WO2021161908A1

WO2021161908A1 - Information processing device and information processing method

Info

Publication number: WO2021161908A1
Application number: PCT/JP2021/004278
Authority: WO
Inventors: 淳也小野
Original assignee: ソニーグループ株式会社
Priority date: 2020-02-14
Filing date: 2021-02-05
Publication date: 2021-08-19
Also published as: JP2021128629A

Abstract

An information processing device according to the present invention is provided with: an execution unit that executes a meaning analysis process for one or more pieces of character information corresponding to one or more languages including a target language which corresponds to utterance by a user; and a calculation unit that calculates an accuracy index value enabling the accuracy of the meaning analysis process to be compared among the languages for each of the one or more pieces of character information on the basis of the result of the meaning analysis process corresponding to each of the one or more pieces of character information.

Description

Information processing device and information processing method

This disclosure relates to an information processing device and an information processing method.

In recent years, technologies related to natural language processing such as semantic analysis and translation have been known. For example, a machine translation system that converts a translation source language sentence into a translation destination language by using a semantic analysis technique is provided.

Japanese Unexamined Patent Publication No. 2004-318344

According to the prior art, the translation result sentence is selected by comparing the result of the semantic analysis of the translation source language sentence with the result of the semantic analysis of the translated translation result sentence.

However, in the prior art, the translation result sentence is selected based on the similarity between the result of the translation source language and the result of the translation destination language. In this case, for example, when the accuracy of the semantic analysis of the translation source language is low, the translation result sentence is selected based on the result of the translation source language with low accuracy, and it is difficult to perform appropriate processing. Further, when there are results of semantic analysis in a plurality of languages, there is also a problem that it is difficult to determine which of the results of the semantic analysis is suitable as the result of the semantic analysis. Therefore, it is desired to enable comparison of semantic analysis accuracy between languages.

Therefore, in this disclosure, we propose an information processing device and an information processing method that can compare the accuracy of semantic analysis between languages.

In order to solve the above problems, the information processing apparatus of one form according to the present disclosure relates to one or more character information corresponding to each of one or more languages including a target language which is a language corresponding to a user's utterance. Based on the execution unit that executes the semantic analysis process and the result of the semantic analysis process corresponding to each of the one or more character information, the accuracy of the semantic analysis process is applied to each of the one or more character information. It is provided with a calculation unit for calculating an accuracy index value that enables comparison between a plurality of languages.

It is a figure which shows an example of information processing which concerns on embodiment of this disclosure. It is a figure which shows the structural example of the information processing system which concerns on embodiment of this disclosure. It is a figure which shows the structural example of the information processing apparatus which concerns on embodiment of this disclosure. It is a figure which shows an example of the language information storage part which concerns on embodiment of this disclosure. It is a figure which shows an example of the semantic frame information storage part which concerns on embodiment of this disclosure. It is a figure which shows an example of the analysis accuracy information storage part which concerns on embodiment of this disclosure. It is a figure which shows an example of the threshold information storage part which concerns on embodiment of this disclosure. It is a figure which shows the structural example of the terminal apparatus which concerns on embodiment of this disclosure. It is a figure which shows an example of the response which concerns on embodiment of this disclosure. It is a flowchart which shows the process of the information processing apparatus which concerns on embodiment of this disclosure. It is a flowchart which shows the process of the information processing system which concerns on embodiment of this disclosure. It is a flowchart which shows the process of the information processing system which concerns on embodiment of this disclosure. It is a flowchart which shows the process of the information processing system which concerns on embodiment of this disclosure. It is a conceptual diagram which shows an example of processing by an information processing system. It is a figure which shows an example of the specific processing by an information processing system. It is a figure which shows an example of analysis accuracy. It is a figure which shows an example of analysis accuracy. It is a figure which shows an example of the relationship between analysis accuracy and score. It is a figure which shows an example of translation through other languages. It is a hardware block diagram which shows an example of the computer which realizes the function of an information processing apparatus.

Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. The information processing apparatus and information processing method according to the present application are not limited by this embodiment. Further, in each of the following embodiments, duplicate description will be omitted by assigning the same reference numerals to the same parts.

The present disclosure will be described according to the order of items shown below.
1. 1. Embodiment 1-1. Outline of information processing according to the embodiment of the present disclosure 1-1-1. Outline, background, effects, etc. 1-1-2. Score function 1-1-2-1. Function generation example 1-1-3. Main flow of processing 1-2. Configuration of Information Processing System According to Embodiment 1-3. Configuration of Information Processing Device According to Embodiment 1-4. Configuration of the terminal device according to the embodiment 1-5. Response example 1-6. Information processing procedure according to the embodiment 1-6-1. Procedure for processing related to information processing equipment 1-6-2. Procedure for processing related to information processing system 1-6-3. Other Procedures for Processing Related to Information Processing Systems Part 1
1-6-4. Other procedure of processing related to information processing system Part 2
1-7. Conceptual diagram of processing by information processing system 1-7-1. Specific examples of processing by an information processing system 1-8. Details of processing, etc. 1-8-1. Language identification 1-8-2. Improvement of low-precision semantic analyzer 1-8-3. Quality estimation 1-8-4. Cloud worker 1-8-5. Translation accuracy and time adjustment 2. Other Embodiments 2-1. Configuration example of performing semantic analysis processing on the client side 2-2. Other configuration examples 2-3. Others 3. Effect of this disclosure 4. Hardware configuration

[1. Embodiment]
[1-1. Outline of information processing according to the embodiment of the present disclosure]
FIG. 1 is a diagram showing an example of information processing according to the embodiment of the present disclosure. The information processing according to the embodiment of the present disclosure is realized by the information processing system 1 (see FIG. 2) including the information processing device 100 (see FIG. 3) and the terminal device 10 (see FIG. 8). FIG. 1 describes an outline of information processing realized by the information processing system 1. FIG. 1 is a diagram showing an example of information processing according to the embodiment of the present disclosure.

First, the terms used in the explanation will be described prior to the explanation in FIG. In the following, the language (input language) corresponding to the user's utterance is described as "target language". For example, the target language corresponds to a language accepted as input by the information processing system 1. In addition, the language to be translated into the target language is described as "translation destination language". That is, one language may be the target language, or the translation destination language for another language may be. For example, if one language is a language corresponding to the user's utterance, it is the target language, and if that one language is the translation destination language of another language, it is the translation destination language. That is, the target language and the translation destination language referred to here are names for distinguishing and expressing languages based on the relationship with other languages of each language in the processing described later. Further, in the following, the character information corresponding to the target language may be described as "input language character information", and the character information corresponding to the translation destination language may be described as "translated sentence" or "translated character information". In addition, among the languages, a language capable of semantic analysis processing is described as a "specific language". In addition, among the languages, a language other than a specific language, that is, a language that cannot perform semantic analysis processing may be described as a "non-specific language".

The outline of the processing will be explained below by the information processing system 1, but the preconditions for the processing will be briefly explained. The information processing system 1 is capable of voice recognition (language identification) and translation processing for many languages, and can accept many languages as input languages (target languages). For example, as shown in FIG. 1, the information processing system 1 can speak many languages such as English, Chinese, Hindi, Spanish, French, Arabic, Portuguese, Bengali, German, Japanese, and Korean. It can be accepted as an input language (target language). Note that these languages are only examples, and the information processing system 1 can accept not only the above languages but also a large number of languages as input languages (target languages).

Further, each process shown in FIG. 1 may be performed by either the information processing device 100 or the terminal device 10 of the information processing system 1. Any device included in the information processing system 1 may perform the processing in which the information processing system 1 is described as the main body of the processing. In the following, a case where the information processing device 100 executes processing such as voice recognition, translation, and semantic analysis in response to the user's utterance detected by the terminal device 10 will be described as an example. In FIG. 1, a case where the information processing device 100 performs processing (information processing) such as voice recognition, translation, and semantic analysis will be described as an example, but even if the terminal device 10 performs these processing (information processing). good. This point will be described later.

From here, FIG. 1 will be specifically described. First, the user speaks. In the example of FIG. 1, a case where the user speaks in Korean is shown. For example, the user utters "Tell me the weather in Tokyo tomorrow (Korean)" in Korean. In this way, when "... (Korean)" is described, it is the language in which the specification is described (for example, Japanese), but in reality, it is assumed that the utterance is in Korean or Korean characters. .. As a result, the information processing system 1 accepts the user's utterance in Korean. For example, the information processing system 1 acquires voice information of utterances in Korean, which is an input language (target language).

Then, the information processing system 1 executes the processing related to voice recognition as shown in the processing phase FS1. The information processing system 1 performs voice recognition processing for voice information spoken by the user. For example, the information processing system 1 acquires the text (character information) of the utterance by the user in the input language (target language) as the utterance information by voice recognition. In the example of FIG. 1, the information processing system 1 determines that the user's utterance is Korean by the language identification process. The information processing system 1 uses character information (speech information) in Korean, which is the target language, as input language character information.

Then, the information processing system 1 develops the utterance sentence (step S1). The information processing system 1 develops an utterance sentence that generates character information (also referred to as "paraphrase") in which the input language character information of the input language "Korean" is paraphrased into another expression while keeping the input language "Korean". .. For example, a paraphrase is character information in which a certain word (character string) in a certain character information is replaced with another word (character string), character information in which a new word (character string) is added to a certain character information, or a certain character. It may be character information generated by various conversion modes such as character information in which information is paraphrased into another expression. For example, the paraphrase may be any character information as long as it is character information converted into another expression while including the meaning of one character information. The information processing system 1 generates paraphrases by appropriately using various techniques related to paraphrases. For example, the information processing system 1 may generate a paraphrase by using information indicating a paraphrase generation rule (paraphrase generation rule information). For example, the information processing device 100 may generate a paraphrase by using the paraphrase generation rule information stored in the storage unit 120 (see FIG. 3). For example, the paraphrase generation rule information may be list information such as a rule for converting a character string such as a flexion of a word ending or a rule for adding a specific character string (a word ending, an object, etc.) to a sentence. Further, the information processing system 1 may transmit the input language character information to the service providing device that provides the paraphrase of a certain character information, and acquire the paraphrase of the input language character information from the service providing device.

The information processing system 1 uses a plurality of paraphrases that paraphrase the input language character information (referred to as "first utterance development sentence") "Tell me the weather in Tokyo tomorrow (Korean)" by expanding the utterance sentence. Generate. For example, the information processing system 1 generates "Tell me the weather in Tokyo tomorrow (Korean)" as a paraphrase of input language character information (referred to as "second utterance development sentence") by utterance sentence expansion. .. For example, the information processing system 1 generates "What is the weather in Tokyo tomorrow? (Korean)" as a paraphrase of input language character information (referred to as "third utterance development sentence") by utterance sentence expansion. do. For example, the information processing system 1 generates "I want to know the weather in Tokyo tomorrow (Korean)" as a paraphrase of input language character information (referred to as "fourth utterance development sentence") by expanding the utterance sentence. .. The above is shown as an example of a paraphrase with a large change for explanation, and the information processing system 1 generates a paraphrase of input language character information by appropriately using various techniques.

In the example of FIG. 1, the information processing system 1 utterably expands the input language character information "Tell me the weather in Tokyo tomorrow (Korean)" into N sentences by the utterance sentence expansion as described above. .. That is, the information processing system 1 generates N sentences including "Tell me the weather in Tokyo tomorrow (Korean)" which is input language character information by expanding the spoken sentences. For example, the information processing system 1 generates a pseudo utterance text list (hereinafter, also referred to as “utterance sentence list”) including input language character information and a paraphrase of the input language character information. The information processing system 1 generates an utterance sentence list including N sentences. In the example of FIG. 1, the information processing system 1 generates an utterance sentence list including N sentences such as a first utterance development sentence, a second utterance development sentence, a third utterance development sentence, and a fourth utterance development sentence.

Then, the information processing system 1 translates each sentence in the utterance sentence list (step S2). The information processing system 1 converts each sentence in the utterance sentence list into a specific language. The information processing system 1 converts each of the N Korean sentences in the utterance sentence list into a specific language. Here, in the example of FIG. 1, M languages such as English, Chinese, Hindi, Spanish, French, Arabic, Portuguese, Bengali, German, Japanese, and Korean are specific languages. Show the case. Note that these languages are only examples, and the information processing system 1 may use a language other than the above languages as a specific language as long as it can perform semantic analysis processing.

The information processing system 1 is a language (specific language) capable of performing semantic analysis processing for each of N sentences such as the first utterance development sentence, the second utterance development sentence, the third utterance development sentence, and the fourth utterance development sentence. Translate to. The information processing system 1 uses English, Chinese, Hindi, Spanish, French, Arabic, Portuguese, Bengali, and Germany for each of the N sentences such as the first utterance development sentence to the fourth utterance development sentence. Translate into M languages (specific languages) such as Chinese, Japanese, and Korean. Hereinafter, generating a sentence of a specific language from a sentence of an input language is also referred to as "language expansion". In FIG. 1, since Korean, which is an input language (target language), is also included in the specific language, the information processing system 1 has the first utterance development sentence, the second utterance development sentence, and the third utterance development sentence for Korean. N sentences such as the expanded sentence and the fourth utterance expanded sentence are used as they are. Further, when the input language (target language) is not included in the specific language, the information processing system 1 does not use the sentence of the input language (speech expansion sentence) for the processing of semantic analysis or the like.

The information processing system 1 has M-1 languages other than Korean among M specific languages, such as a first utterance development sentence, a second utterance development sentence, a third utterance development sentence, and a fourth utterance development sentence. Translate each of the N sentences. For example, the information processing system 1 translates each of N sentences such as the first utterance development sentence, the second utterance development sentence, the third utterance development sentence, and the fourth utterance development sentence into English.

For example, the information processing system 1 converts the first utterance development sentence "Tell me the weather in Tokyo tomorrow (Korean)" into a translated sentence "Tell me the weather in Tokyo tomorrow" corresponding to English. For example, the information processing system 1 converts the second utterance development sentence "Tell me the weather in Tokyo tomorrow (Korean)" into the translated sentence "Please tell me the weather in Tokyo tomorrow" corresponding to English. For example, the information processing system 1 converts the third utterance development sentence "What is the weather in Tokyo tomorrow? (Korean)" into a translation sentence "What is the weather in Tokyo tomorrow?" Corresponding to English. .. For example, the information information system 1 converts the fourth utterance development sentence "I want to know the weather in Tokyo tomorrow (Korean)" into a translation sentence "I want to know the weather of Tokyo tomorrow" corresponding to English. .. In this way, the information processing system 1 translates N English sentences corresponding to each of the N sentences such as the first utterance development sentence, the second utterance development sentence, the third utterance development sentence, and the fourth utterance development sentence. Generate a statement.

Similarly, the information processing system 1 translates N Chinese characters corresponding to each of N sentences such as the first utterance development sentence, the second utterance development sentence, the third utterance development sentence, and the fourth utterance development sentence. Generate a statement. Further, the information processing system 1 has N translations of N Hindi corresponding to each of N sentences such as the first utterance development sentence, the second utterance development sentence, the third utterance development sentence, and the fourth utterance development sentence. To generate. In addition, the information processing system 1 also applies to Spanish, French, Arabic, Portuguese, Bengali, German, Japanese, etc. in the same manner as the first utterance development sentence, the second utterance development sentence, and the third utterance development sentence. , N translation sentences corresponding to each of N sentences such as the fourth utterance expansion sentence are generated for each of M languages.

By the above-mentioned processing, the information processing system 1 generates N × M sentences including N sentences in Korean and N translated sentences in each of M-1 languages. The information processing system 1 generates a list (also referred to as a "translation list") including N sentences in Korean and N translations of each of M-1 languages. The information processing system 1 generates a translation list containing N × M sentences, which are N sentences in each of the M languages. In this way, the information processing system 1 executes the processing of the utterance sentence expansion and the language expansion in the processing phase FS1. As described above, the processing phase FS1 is a processing that depends on the input language.

Then, the information processing system 1 executes the processing related to the semantic analysis as shown in the processing phase FS2. The information processing system 1 performs an utterance semantic analysis process (also referred to as “semantic analysis process”) (step S3). The information processing system 1 performs a semantic analysis process using each sentence included in the translation list (also referred to as a "semantic analysis target sentence"). The information processing system 1 generates information on the semantic frame as an analysis result by the semantic analysis process. In the example of FIG. 1, the information processing system 1 generates information of N × M semantic frames corresponding to each of N × M sentences. The information processing system 1 may generate more than N × M semantic frames. Although FIG. 1 shows a case where one meaning frame is generated for one sentence for explanation, the information processing system 1 may generate a plurality of meaning frames for one sentence. The information processing system 1 generates information on a semantic frame including a score (also referred to as a “semantic analysis score”) indicating the accuracy of the semantic analysis by the semantic analysis process. The information processing system 1 uses a semantic analyzer corresponding to the language of the sentence to generate information on a semantic frame including a semantic analysis score indicating the accuracy of the semantic analysis of the sentence.

For example, the information processing system 1 uses an English semantic analyzer to perform semantic analysis processing on N English sentences to indicate the accuracy of the semantic analysis of each of the N English sentences. Generates information on N semantic frames including analysis scores. For example, the information processing system 1 inputs an English translation "Tell me the weather in Tokyo tomorrow" (also referred to as a "first semantic analysis target sentence") corresponding to the first utterance development sentence into an English semantic analyzer. By doing so, the information of the semantic frame including the semantic analysis score is generated. The information processing system 1 identifies the Domain-Goal of the first semantic analysis target sentence as "Weather-Check". Further, in the information processing system 1, the slot value (also referred to as "Value") of the Attribute "Date" corresponding to the Domain-Goal "Weather-Check" is "tomorrow", and the slot value of the Attribute "Place" is "Tokyo". Is presumed to be. As a result, in the information processing system 1, the domain-Goal is "Weather-Check", the slot value of the Attribute "Date" is "tomorrow", and the slot value of the Attribute "Place" is "Tokyo". Generates the information of the meaning frame to be shown. Further, the information processing system 1 obtains a score "0.9" indicating the certainty of the semantic analysis output by the English semantic analyzer into which the first semantic analysis target sentence is input, and performs a semantic analysis of the first semantic analysis target sentence. Used as a score.

In addition, the information processing system 1 also uses an English semantic analyzer for the remaining N-1 translated sentences, and N-1 semantic frames including the semantic analysis scores of each of the N-1 sentences. Generate information about. As a result, the information processing system 1 generates information of N semantic frames including the semantic analysis score of each of the N sentences in English.

Further, the information processing system 1 also uses the semantic analyzer of each language for each of the M-1 languages other than English to generate information of the semantic frame including the semantic analysis score. For example, the information processing system 1 uses a Korean semantic analyzer to generate information on N semantic frames including the semantic analysis score of each of the N sentences in Korean. The information processing system 1 uses a Japanese semantic analyzer to generate information on N semantic frames including the semantic analysis score of each of the N sentences in Japanese. Similarly, the information processing system 1 also uses a semantic analyzer of each language for Chinese, Hindi, Spanish, French, Arabic, Portuguese, Bengali, German, etc., and N pieces of each language. Generates information about the semantic frame containing the semantic analysis score for each of the sentences in.

As a result, the information processing system 1 generates information of N × M semantic frames including the semantic analysis score of N × M semantic analysis target sentences. The information processing system 1 does not have to generate the information of the semantic frame for all the N × M semantic analysis target sentences. For example, the information processing system 1 may perform the semantic analysis process except for the sentences having low translation quality among the N × M semantic analysis target sentences, and the details of this point will be described later.

Here, the details will be described later, but the semantic analyzer (function of semantic analysis processing) of each language is created for each language, and the relationship between the languages is not considered. Therefore, the semantic analysis score generated by the semantic analyzer (function of semantic analysis processing) of each language indicates the certainty (accuracy) within that language, and the score considering the relationship with other languages No. Therefore, although the accuracy of such a semantic analysis score tends to be higher as the value is larger, when it is used as it is for the comparison of the accuracy of the semantic analysis between languages, it may not be an appropriate comparison. For example, the Japanese semantic analysis score "0.9" and the English semantic analysis score "0.9" do not always show the same accuracy.

Therefore, the information processing system 1 performs a process of converting the semantic analysis score into an accuracy index value that enables comparison between a plurality of languages. The information processing system 1 performs a process of converting the semantic analysis score into the semantic analysis accuracy (%), which is an accuracy index value that enables comparison between a plurality of languages (step S4). The information processing system 1 calculates the semantic analysis accuracy (%) using the semantic analysis score. In FIG. 1, the case where the accuracy index value "semantic analysis accuracy (%)" is used is shown as an example, but the accuracy index value is "semantic analysis accuracy (semantic analysis accuracy (%)" if the semantic analysis accuracy between languages can be compared. %) ”, Any information may be used. Further, the unit is not limited to "%", and may be various units or may have no unit.

The information processing system 1 calculates the semantic analysis accuracy by using a function as shown in the following equation (1) (hereinafter, also referred to as a “score function”). For example, the information processing apparatus 100 calculates the semantic analysis accuracy by using the score function stored in the storage unit 120. The details of the generation of the score function that outputs the accuracy index value that enables the accuracy of the semantic analysis process to be compared between a plurality of languages will be described later.

Acc = f (lang, score)

"Acc" in equation (1) indicates the accuracy of semantic analysis. Further, "score" in the formula (1) indicates a semantic analysis score. In "score", the semantic analysis score to be converted into the semantic analysis accuracy is input. "Lang" in the formula (1) indicates a language. In "lang", information indicating the language that is the target of the semantic analysis score is input. “F ()” in the equation (1) indicates a score function that outputs the semantic analysis accuracy by inputting the language specification and the semantic analysis score.

In this way, the information processing system 1 uses the equation (1) to convert the semantic analysis score so that it can be compared between a plurality of languages according to the semantic analysis score and the language designation (meaning). Analysis accuracy) is calculated. The equation (1) is an example, and the information processing system 1 may perform a process of calculating the semantic analysis accuracy by using a function other than the equation (1).

For example, the information processing system 1 calculates the semantic analysis accuracy by using a function for each language (score function for each language). In this case, the information processing system 1 calculates the semantic analysis accuracy by using the language-specific score function that outputs the semantic analysis accuracy in response to the input of the semantic analysis score. For example, the information processing apparatus 100 calculates the semantic analysis accuracy by using the language-specific score function stored in the storage unit 120. The information processing system 1 selects a language-specific score function corresponding to the language of the semantic analysis score from a plurality of language-specific score functions, and calculates the semantic analysis accuracy using the selected language-specific score function.

For example, when calculating the semantic analysis accuracy for English, the information processing system 1 calculates the semantic analysis accuracy by using the English score function, which is a language-specific score function for English. For example, when the information processing system 1 calculates the semantic analysis accuracy for Japanese, the semantic analysis accuracy is calculated by using the Japanese score function, which is a language-specific score function for Japanese.

Further, the function of Eq. (1) may be a program including conditional branching by language. In this case, the function of the equation (1) may include a conditional branch according to the comparison between the variable "lang" and each language. For example, the function of the equation (1) may be a function (program) that calculates the semantic analysis accuracy using the language-specific score function of the language in which the variable “lang” matches and outputs the result. As described above, the function of the equation (1) may be a function (program) including a plurality of language-specific score functions corresponding to each language and a conditional branch of which language-specific score function is used.

The information processing system 1 uses the equation (1) to convert each of the semantic analysis scores of N × M semantic analysis target sentences into the semantic analysis accuracy. The information processing system 1 calculates the semantic analysis accuracy of N × M semantic analysis target sentences by using the equation (1). For example, the information processing system 1 uses the semantic analysis score "0.9" of the first semantic analysis target sentence and the information indicating the language "English" of the first semantic analysis target sentence, and uses the information indicating the language "English" of the first semantic analysis target sentence. Calculate the semantic analysis accuracy of. The information processing system 1 inputs the semantic analysis score "0.9" of the first semantic analysis target sentence and the information indicating the language "English" of the first semantic analysis target sentence into the equation (1). 1 Semantic analysis Calculate the semantic analysis accuracy of the target sentence. In the example of FIG. 1, the information processing system 1 calculates the semantic analysis accuracy of the first semantic analysis target sentence as "96.6 (%)". Similarly, the information processing system 1 calculates the semantic analysis accuracy for N × M-1 sentences (semantic analysis target sentences) other than the first semantic analysis target sentence. As a result, the information processing system 1 calculates the semantic analysis accuracy of all N × M sentences (semantic analysis target sentences) in the translation list. In this way, the information processing system 1 executes the semantic analysis process on the sentences (N × M sentences) in the translation list in the processing phase FS2, and calculates the semantic analysis accuracy of the N × M sentences. .. As described above, the processing phase FS2 is an input language-independent process.

Then, the information processing system 1 executes the processing related to the response generation as shown in the processing phase FS3. In FIG. 1, the information processing system 1 selects a semantic frame (semantic analysis target sentence) to be used for response generation or the like prior to response generation. The information processing system 1 selects a sentence (also referred to as “character information to be processed”) used for response generation or the like from the sentences in the translation list. The information processing system 1 selects a meaning frame used for response generation or the like from N × M meaning frames. The information processing system 1 selects a semantic frame (semantic analysis target sentence) having the maximum semantic analysis accuracy from N × M semantic frames. In the example of FIG. 1, the information processing system 1 selects the first semantic analysis target sentence having a semantic analysis accuracy of "96.6 (%)" as the semantic frame (semantic analysis target sentence) used for response generation or the like. The information processing system 1 has a semantic frame included in the semantic analysis result of the English translation "Tell me the weather in Tokyo tomorrow" whose semantic analysis accuracy is "96.6 (%)" (also referred to as "processed semantic frame"). Is selected as the information used for response generation, etc.

Then, the information processing system 1 performs slot inverse transformation (step S5). The information processing system 1 converts the slot value in the processing target meaning frame into the slot value of the input language (target language). The information processing system 1 converts the slot value of the specific language (translation destination language) into the slot value of the input language (target language).

In FIG. 1, the information processing system 1 converts the slot value of English, which is the translation destination language, into the slot value of Korean, which is the input language. For example, the information processing system 1 converts the slot value of the Attribute “Date” from “tomorrow” to “tomorrow (Korean)” and converts the slot value from English to Korean. For example, the information processing system 1 converts the slot value of the Attribute “Place” from “Tokyo” to “Tokyo (Korean)” and converts the slot value from English to Korean.

Then, the information processing system 1 determines the service to be started (step S6). For example, the information processing system 1 determines a service to be started from various services such as a calendar service SV1, a weather service SV2, an alarm service SV3, and a music service SV4. In FIG. 1, since the user is asking for the weather, the information processing system 1 decides to start the weather service SV2. Then, the information processing system 1 generates a response. For example, the information processing system 1 outputs information indicating the weather in Tokyo tomorrow in Korean. For example, the information processing system 1 outputs a response such as "Tomorrow's weather in Tokyo is sunny (Korean)" by voice in Korean or displays it in Korean. As described above, the processing phase FS3 is a processing that depends on the input language.

As described above, in FIG. 1, the information processing system 1 performs semantic analysis processing for each of a plurality of sentences in a plurality of languages generated by the processing of utterance sentence expansion and language expansion. Then, the information processing system 1 converts the semantic analysis score generated by the processing of the semantic analysis into an accuracy index value (semantic analysis accuracy) that enables comparison between a plurality of languages. As a result, the information processing system 1 can compare the accuracy of semantic analysis between languages.

[1-1-1. Overview, background and effects]
As described above, the information processing system 1 accepts a user's utterance to a device such as a smart speaker, generates a table structure of meaning frames from the utterance content, classifies the domain goals of the utterance, and takes out slots. As a result, the information processing system 1 provides a mechanism for improving the accuracy of the semantic analysis process for extracting information necessary for executing an application or service.

As shown in FIG. 1, the method executed by the information processing system 1 generates an utterance sentence list in which phrases and phrases are increased in variations by different expressions by expanding the utterance sentences for the utterance sentences after speech recognition. do. Further, the method executed by the information processing system 1 performs language conversion by translating the utterance sentence list into all specific languages that can be processed by the utterance meaning analyzer. As a result, the information processing system 1 can make a judgment from various expressions and languages, although normally only the semantic analysis result of only the input language can be obtained.

As described above, even in the case of an input language in which the analysis accuracy of the semantic analyzer is not sufficient, the information processing system 1 has sufficient analysis accuracy by translating into all specific languages that can be processed by the speech semantic analyzer. It enables analysis in language. In addition, since the translation accuracy of the translator in language conversion is greatly affected, the information processing system 1 maintains the same meaning by expanding the utterance sentence, and does not affect the result of the meaning frame, so that the character string on the surface layer is not affected. Absorb by increasing the expression. The information processing system 1 prepares a module for estimating the quality of translation accuracy because the semantic analysis results increase due to utterance sentence expansion and language expansion, and utterances with accuracy below a certain level may be rejected. Will be described later. In addition, the information processing system 1 prepares in advance a corresponding function (score function) between the score value of the semantic analysis process and the percentage of the analysis accuracy for each language, so that the normal score value does not make sense for comparison between languages. It is possible to compare the semantic analysis results in the above from the viewpoint of analysis accuracy.

In order to create a semantic analysis process (semantic analyzer, etc.), it is necessary to collect utterance sentences according to the domain goal, and from there, it is necessary to perform labeling to cut out as a slot. Therefore, in terms of understanding the target language, designing the standard for domain goals, and collecting and labeling the corpus, localizing it into one language is compared to machine translation, where bilingual collection is the main cost. The load is high, and multilingual development of semantic analysis generally requires time and labor costs. Localization is an unavoidable issue when expanding our business globally. In addition, if the utterance sentence collection is insufficient or if there is a language in which the semantic frame design is not sufficient, it is difficult to obtain sufficient semantic analysis results for that language.

On the other hand, the information processing system 1 has the following technical features. The information processing system 1 improves or supports the accuracy of semantic analysis by using translation technology. Further, the information processing system 1 performs utterance development that converts the input utterance sentence into an expression having the same meaning, for example, an expression that is not affected by the output result of the meaning frame. Further, the information processing system 1 translates the increased utterance sentences due to the utterance development into a language that can be processed by the semantic analyzer.

In addition, the information processing system 1 obtains a semantic frame, which is the result of semantic analysis, for each utterance sentence that has been linguistically converted by a translator for each number of expansions x number of languages. Further, the information processing system 1 calculates the score of the semantic analysis result and the accuracy index of the percentage display from the evaluation data in advance, and converts it into a score accuracy correspondence table or a function. Further, the information processing system 1 generates an accuracy function (score function for each language, etc.) with accuracy = f (score).

In addition, the information processing system 1 passes the accuracy function from the score of the semantic analysis for each combination of language and expansion, and obtains the one with the maximum accuracy as the semantic analysis result. Further, in order to return a response in the input language, the information processing system 1 reversely converts the slot value of the semantic frame from the language (specific language) for which the semantic analysis corresponding to the semantic frame is performed to the input language.

The information processing system 1 outputs in a format that can be understood which language it is in three phases of input, analysis, and output when responding. For example, the information processing system 1 outputs in the form of an image (icon or the like), a voice (sound effect or the like), a text (language name / language code) or the like. The details of this point will be described later.

Further, the information processing system 1 interrupts the processing when the translation accuracy and the semantic analysis accuracy are below a certain level, and presents the reason for the interruption. Executing an application or service with low accuracy often does not produce the results that the user expects. Therefore, the information processing system 1 tells the user whether the translation process was not successful or the translation was successful but the semantic analysis process was not successful, so that the user can adjust (control) the next time the input is performed. Allows you to.

Here, an example of conventional translation will be briefly described with reference to FIG. FIG. 19 is a diagram showing an example of translation via another language. FIG. 19 shows an example of a method via a specific language in multilingualization.

In translation technology, a large number of pairs of bilingual sentences are required to train a translation model, and especially when translating from minor language A to minor language B, it becomes difficult to collect a corpus, and the translation engine. Will be difficult to realize. The minor language referred to here means a language in which it is difficult to collect a sufficient amount of data, for example, in the country to which the provider of the information processing system 1 belongs. On the other hand, the major language means a language in which it is relatively easy to collect a sufficient amount of data, for example, in the country to which the provider of the information processing system 1 belongs. For example, the major language includes a language used in the country to which the provider of the information processing system 1 belongs. A minor language may be a language in which the number of people (speakers) who use the language is relatively small, and a major language is a language in which the number of people (speakers) who use the language is relatively large. It may be.

Therefore, as shown in Fig. 19, there is a method to convert to a major language once. In the example of FIG. 19, after the minor language A is translated into the major language, the major language is translated into the minor language B twice. There are two reasons for translating twice in this way. The first is that the cost from minor language A to major language and from major language to minor language B is lower than the cost required to collect and create bilingual sentences from minor language A to minor language B. Secondly, since the major language is a major language, there is a need for multilingual support from minor language A to major language, and from major language to minor language B, and in many cases it has already been supported and can be translated. This is because there is a high possibility that the existing translator (translation processing) can be used as it is (it can be diverted).

As mentioned above, a method of translating into a specific language (for example, a major language) and then retranslating into the target language can be considered. The above is a case where translators having the same purpose are simply combined (single-bonded), and since they have the same properties and purposes, there is a high possibility that they will function well.

On the other hand, the object in the example shown in FIG. 1 is that a multilingual translator is used to improve the accuracy of the semantic analyzer. However, when translating text after speech recognition into a specific language, the translator and the semantic analyzer have different properties and purposes, and may not function well with a single bond. In addition, it is difficult to compare and select the semantic analysis results for the translated sentences converted into each language. The reasons for this are described below as problems with the prior art.

In general, translators tend to have a literary tone of written words, which aims to be widely used by translators. This is due to the fact that it is often learned based on a corpus that is not ambiguous and has a high degree of perfection as a sentence. On the other hand, in the case of a style in which a person asks a machine, such as a smart speaker or an AI (Artificial Intelligence) chatbot, or a style in which a person communicates with a machine, the expression of words such as a person-to-person conversation is naturally used. Has been done.

That is, the following two points can cause the translation process (translator) and the semantic analysis process (semantic analyzer) to not function well when simply combined (single bond).

The first point is that, for example, the input information to be input is not the written language but the spoken language. The second point is that, for example, an appropriate translation cannot be obtained for a language-specific expression. However, recent speech translators may also incorporate (correspond to) spoken language (colloquial tone), and it is important to address the second point below.

Regarding the second point, there is no guarantee that the expression of the verb "kake" such as "play music" will be translated correctly through the translator. Moreover, even if the translation is properly performed when viewed as a single translator, it may be an unexpected expression when viewed from the semantic analyzer. This is because it depends on the training sentences for each domain goal in the learning phase. There is a gap between the accuracy of the translator and the accuracy of the semantic analyzer.

On the other hand, as described above, the information processing system 1 expands the user's utterance into a plurality of expressions (plural utterance expansion sentences) by expanding the utterance sentence. Then, the information processing system 1 uses a translator of a plurality of languages to develop each of the plurality of utterance development sentences into a language. Then, the information processing system 1 calculates an accuracy index value for making it possible to compare the semantic analysis accuracy between a plurality of languages, and using the calculated accuracy index value, which language has the better semantic analysis accuracy. Can be compared. The information processing system 1 performs processing such as response generation using a language that can be processed accurately among each language. As a result, the information processing system 1 selects a sentence that is expected to have good accuracy from a large number of candidates even if there is a gap between the accuracy of the translator and the accuracy of the semantic analyzer. By processing, response processing and the like can be performed based on accurate semantic analysis.

[1-1-2. Score function]
From here, the score function will be described. As described above, the score function of the equation (1) is a function that converts the analysis accuracy into a percentage from the score of the semantic analysis result (semantic analysis score) and the analysis language.

For example, semantic analysis is learned and modeled based on the corpus collected for each language. That is, the learning is not performed assuming the inter-language at the time of learning. Therefore, the score value indicating the statistical superiority at the time of inference can be relatively used in the language of the learning model, but the score value of the language A and the language B, etc. beyond the language (semantic analysis). It doesn't make sense to compare scores). For example, regarding the score (semantic analysis score), when English is "0.4" and German is "0.5", there is no reason to select German with a high numerical value.

Therefore, the information processing system 1 calculates an index value (semantic analysis accuracy) that enables accuracy comparison between languages using a score function as in Eq. (1).

[1-1-2-1. Function generation example]
Hereinafter, an example of generating a score function will be described with reference to FIGS. 16 to 18. 16 and 17 are diagrams showing an example of analysis accuracy. FIG. 16 is a diagram showing an example of Japanese analysis accuracy. FIG. 17 is a diagram showing an example of English analysis accuracy. FIG. 18 is a diagram showing an example of the relationship between the analysis accuracy and the score.

For example, the information processing system 1 calculates the probability of correct answer (percentage of analysis accuracy) when collecting the results of a value equal to or higher than the score of a certain value from the evaluation data in the language, changes a certain value, and evaluates again. Calculate the probability of correct answer. As a result, the information processing system 1 can create a score table for each language as shown in FIGS. 16 and 17.

For example, the information processing system 1 calculates the probability of a correct answer (percentage of analysis accuracy) when collecting results of a value equal to or higher than a certain value for Japanese, changes a certain value, evaluates it again, and corrects the answer. By calculating the probability of, a score table as shown in FIG. 16 is generated. The information processing system 1 calculates the probability of a correct answer (percentage of analysis accuracy) when collecting results of a value equal to or higher than a certain value for English, changes a certain value, evaluates it again, and determines the probability of the correct answer. By calculating, a score table as shown in FIG. 17 is generated. The information processing system 1 generates a score table for each language (specific language) by performing the same processing for each language (specific language) such as Chinese and Spanish.

To explain the case of targeting Japanese as an example, the information processing system 1 has evaluation data in which the text (character information) after voice recognition in Japanese and the result of semantic analysis of the character information are associated with each other. To generate a Japanese score table using. For example, the information processing system 1 provides evaluation data in which the text (character information) after voice recognition in Japanese and the result of its semantic analysis are associated with information (label) indicating whether or not the result is correct. Use to generate a Japanese score table. For example, the information processing system 1 has a semantic analysis score indicating the certainty of semantic analysis for each text (character information) and a label indicating whether the specified domain goal or the like for each text (character information) is correct. A Japanese score table is generated using the associated evaluation data. The label corresponding to the result of the semantic analysis of each text (character information) may be set by the administrator of the information processing system 1 or the like, or may be set automatically. Further, the label may be assigned to the result of the semantic analysis of the corresponding text (character information) by the administrator of the information processing system 1, or may be automatically assigned.

The information processing system 1 analyzes the meaning of a text (character information) when the domain goal or slot, which is the result of the semantic analysis with a semantic analysis score of "0.4", is incorrect. Evaluation data in which the score "0.4" is associated with the label "0" indicating an incorrect answer is used. Further, in the information processing system 1, when the domain goal and the slot, which are the results of the semantic analysis in which the semantic analysis score of a certain text (character information) is "0.9", are correct, the meaning of the text (character information). Evaluation data in which the analysis score "0.9" is associated with the label "1" indicating the correct answer is used. The information processing system 1 generates a Japanese score table using the evaluation data in which the semantic analysis score of each text (character information) and the label are associated with each other.

The information processing system 1 generates the graph GR1 as shown in FIG. 18 based on the information showing the relationship between the semantic analysis score of each language and the correct answer rate (semantic analysis accuracy) as shown in FIGS. 16 and 17. .. The information processing system 1 generates a score function (score function for each language) of each language based on the graph GR1 as shown in FIG. In FIG. 18, black circles connected by solid lines show the results in Japanese, and white circles connected by dotted lines show the results in English.

Here, in the graph GR1 of FIG. 18, the horizontal axis indicates the “score threshold”, that is, the threshold of the semantic analysis score, and the vertical axis indicates the “probability of correct answer (unit: percentage)”, that is, the semantic analysis accuracy. As described above, the graph GR1 of FIG. 18 shows the ratio of the analysis results in which the estimated information (domain goal, etc.) is correct in the analysis result group whose semantic analysis score is equal to or higher than the score threshold value for each language. For example, for Japanese, it is shown that 60% of the analysis result group whose semantic analysis score is "0.6" or more is the analysis result in which the estimated information is correct. For example, in the information processing system 1, for Japanese, the number of analysis results having a semantic analysis score of "0.6" or more is "10000", and the number of analysis results whose estimated information is correct is "6000". In the case of ", the probability of correct answer (unit: percentage)", that is, the semantic analysis accuracy is calculated as "60%". For example, where the horizontal axis has a low "score threshold", there is (statistically) variation, and where the horizontal axis has a high "score threshold", the value is stable and high. For example, when the horizontal axis is where the "score threshold" exceeds "0.5", the increase is monotonous.

For example, the information processing system 1 derives (generates) a fitting function in which the horizontal axis is the "score threshold" and the vertical axis is the "probability of correct answer (unit: percentage)". Then, the information processing system 1 uses the fitting function as the score function. The information processing system 1 generates a score function for each language by calculating a fitting function for each language.

Then, the information processing system 1 generates the equation (1) by generating a function (program) using the score function for each language. As a result, the information processing system 1 generates the score function (program) shown in the equation (1) for calculating the semantic analysis accuracy (acc) from the analysis language (lang) and the semantic analysis score (score) of the language.

[1-1-3. Main flow of processing]
The main flow of processing will be briefly described below.

The information processing system 1 performs the following processing for utterance input. The information processing system 1 converts utterances into texts by voice recognition. Here, the input language is predetermined as a language that can be supported by the information processing system 1, such as Spanish.

The information processing system 1 performs the following processing for the automatic expansion of utterance sentences. First, as a premise, the translator and the semantic analyzer have different purposes. The corpus of bilingual sentences and the corpus of each domain goal are collected and created separately, and are learned and modeled separately. Therefore, there is a gap between the translated sentence output from the translator model and the analysis sentence assumed by the semantic analyzer model.

Therefore, the information processing system 1 automatically changes the wording and the phrase of the utterance sentence under the condition that the output of the meaning frame does not change, and generates a pseudo utterance sentence list. For example, the information processing system 1 responds to the input utterance "play music" by "play music", "play music", "play music", "music play", and the like. Generate paraphrases.

The information processing system 1 performs the following processing from the input language to the language development by translation. The information processing system 1 uses the pseudo-utterance sentence list as input to the translator and translates each pseudo-utterance sentence into all languages that can be processed by the semantic analyzer. The information processing system 1 generates N pseudo utterance sentences from one utterance sentence and translates them into M language. In this case, the number of elements of the translated sentence list output from the translator is N × M.

As described above, when the input language is a specific language, the translation sentence list may include sentences in the input language. For example, the information processing system 1 generates N pseudo utterance sentences including the one utterance sentence from one utterance sentence. Then, when the language (input language) of the utterance sentence is a specific language, the information processing system 1 translates it into the M-1 language and generates N sentences for each of the M languages including the input language. In this case, the number of elements in the translated sentence list is N × M.

The information processing system 1 performs the following processing on the semantic analysis processing (speech semantic analyzer). The information processing system 1 performs utterance semantic analysis processing for each language and utterance sentence development. The information processing system 1 generates a semantic frame from a text (character string) by a semantic analysis process. For example, the information processing system 1 is used to determine which application or service to execute by the function of the semantic analysis process (speech semantic analyzer) and to specifically execute the application or service. Generates semantic frame information in tabular form including slot information to be created.

For example, as shown in FIG. 6, the information processing system 1 executes a semantic analysis process on an element of a translated sentence list (in each language for utterance development in FIG. 6), and obtains a semantic frame and an analysis score value. In the example of FIG. 6, the information processing system 1 executes a semantic analysis process on 40 sentences in the translated sentence list, and obtains a semantic frame and an analysis score value. That is, in the example of FIG. 6, the information processing system 1 executes a semantic analysis process on 40 sentences obtained by multiplying the number of utterance development sentences "4" by the number of languages "10", and executes a semantic frame. And get the analysis score value.

The information processing system 1 performs the following processing for conversion from the score value (semantic analysis score) to the analysis accuracy (semantic analysis accuracy). First, the score value (semantic analysis score) expresses the estimated probability, and the higher the value, the more statistically superior it becomes. In other words, the higher the score value (semantic analysis score), the higher the estimation accuracy. Here, in the same language, since it is judged from the same model from the pseudo utterance sentence list, it is possible to select the utterance sentence and the meaning frame having the highest score value.

However, since the model is different for each language, it does not make sense to compare the score values between the languages. Therefore, the information processing system 1 calculates the analysis accuracy (semantic analysis accuracy) as shown in FIG. 6 from the score function by inputting the analysis language and the score value.

The information processing system 1 performs the following processing for analysis language and semantic frame selection. The information processing system 1 selects the result of the analysis language, utterance sentence, and semantic frame having the highest analysis accuracy (semantic analysis accuracy) from the analysis accuracy (semantic analysis accuracy).

The information processing system 1 performs the following processing for the slot inverse transformation of the semantic frame. The semantic frame of the maximum analysis accuracy is the analysis result in a specific language, and the domain goal is a language-independent abstract expression, so it is language-independent. On the other hand, the value of the slot information is expressed in a specific language, and the application or service generally needs to return a response in the input language spoken by the user. Therefore, the information processing system 1 may reversely convert the slot information from the specific language to the input language in order to accurately execute the application or service.

The information processing system 1 performs inverse transformation by the following method. The first method is to convert words / phrases from a knowledge database (DB) using a multilingual conversion dictionary. The second method is to reverse-translate into a specific language with a translator. The information processing system 1 is not limited to the method described above, and may perform inverse transformation by various methods.

Note that the slot information of the meaning frame is basically an artist name, song name, place name, device name, etc., and is a word or phrase rather than a sentence. Can be done. Therefore, the slot information can be accurately converted by the dictionary. On the other hand, in the case of a notation not found in the knowledge DB, the information processing system 1 may perform reverse translation from a specific language to an input language by reverse translation of the second method, and perform two-step conversion.

The information processing system 1 performs the following processing for response generation. For example, in the information processing system 1, when responding, it can be seen that the semantic analysis process (speech semantic analyzer) is not a normal supported language but a pseudo-processed language that is not the target by the translator. Output in a format that shows which language was processed in three phases: input, analysis, and output. An example of this point will be described with reference to FIG. For example, the information processing system 1 outputs in the form of an image (icon or the like), a voice (sound effect or the like), a text (language name / language code) or the like.

[1-2. Configuration of information processing system according to the embodiment]
The information processing system 1 shown in FIG. 2 will be described. As shown in FIG. 2, the information processing system 1 includes a terminal device 10 and an information processing device 100. The terminal device 10 and the information processing device 100 are connected to each other via a predetermined communication network (network N) so as to be communicable by wire or wirelessly. FIG. 2 is a diagram showing a configuration example of an information processing system according to an embodiment. The information processing system 1 shown in FIG. 2 may include a plurality of terminal devices 10 and a plurality of information processing devices 100. For example, the information processing system 1 realizes the above-mentioned dialogue system.

The information processing device 100 is a computer that converts input language character information corresponding to a user's utterance in the target language into a translated sentence of the translation destination language and executes a semantic analysis process. The information processing device 100 performs an inverse transformation process of converting the result of the semantic analysis process corresponding to the translation destination language into the target language. Further, the information processing device 100 is a computer that transmits various information to the terminal device 10. The information processing device 100 is a server device used to provide services related to various functions. For example, the information processing device 100 is used to provide the user with a service related to the dialogue system. The information processing device 100 performs various information processing related to the dialogue system to the user.

Further, the information processing device 100 may have software modules such as voice signal processing, voice recognition, utterance semantic analysis, and dialogue control. For example, the information processing apparatus 100 may have functions of natural language understanding (NLU: Natural Language Understanding) and automatic speech recognition (ASR: Automatic Speech Recognition). For example, the information processing device 100 may estimate information about a user's intent (intention) or entity (target) from input information uttered by the user. The information processing device 100 functions as a voice recognition server having functions of natural language understanding and automatic voice recognition.

The terminal device 10 is a computer that detects the user's utterance and transmits the voice of the user's utterance to the information processing device 100 or the like. Further, the terminal device 10 may have functions such as natural language understanding and automatic voice recognition. For example, the terminal device 10 may estimate information about a user's intent (intention) or entity (target) from input information uttered by the user. The terminal device 10 is a device device used by a user. The terminal device 10 accepts input by the user. The terminal device 10 accepts voice input by the user's utterance and input by the user's operation. The terminal device 10 displays information according to the input of the user.

The terminal device 10 is an information processing device used by the user. The terminal device 10 is used to provide a dialogue service that responds to a user's utterance. The terminal device 10 has a sound sensor that detects the sound of a microphone or the like. For example, the terminal device 10 uses a sound sensor to detect a user's utterance around the terminal device 10. For example, the terminal device 10 may be a device (voice assist terminal) that detects ambient sounds and performs various processes according to the detected sounds. The terminal device 10 is a computer that processes a user's utterance.

The terminal device 10 may be any device as long as the processing in the embodiment can be realized. The terminal device 10 may be any device as long as it has a function of detecting the user's utterance and transmitting it to the information processing device 100. The terminal device 10 may be, for example, a device such as a smartphone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, or a PDA (Personal Digital Assistant). The terminal device 10 may be a wearable terminal (Wearable Device) or the like that the user can wear. For example, the terminal device 10 may be a wristwatch-type terminal, a glasses-type terminal, or the like. Further, the terminal device 10 may be a so-called home electric appliance such as a television or a refrigerator. For example, the terminal device 10 may be a robot that interacts with a human (user), such as a smart speaker, an entertainment robot, or a domestic robot. Further, the terminal device 10 may be a device arranged at a predetermined position such as digital signage.

[1-3. Configuration of Information Processing Device According to Embodiment]
Next, the configuration of the information processing device 100, which is an example of the information processing device that executes the information processing according to the embodiment, will be described. FIG. 3 is a diagram showing a configuration example of the information processing device 100 according to the embodiment of the present disclosure.

As shown in FIG. 3, the information processing device 100 includes a communication unit 110, a storage unit 120, and a control unit 130. The information processing device 100 includes an input unit (for example, a keyboard, a mouse, etc.) that receives various operations from the administrator of the information processing device 100, and a display unit (for example, a liquid crystal display, etc.) for displaying various information. You may have.

The communication unit 110 is realized by, for example, a NIC (Network Interface Card) or the like. Then, the communication unit 110 is connected to the network N (see FIG. 2) by wire or wirelessly, and transmits / receives information to / from another information processing device such as the terminal device 10. Further, the communication unit 110 may send and receive information to and from the terminal device 10.

The storage unit 120 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory (Flash Memory), or a storage device such as a hard disk or an optical disk. As shown in FIG. 3, the storage unit 120 according to the embodiment includes a language information storage unit 121, a semantic frame information storage unit 122, an analysis accuracy information storage unit 123, a threshold information storage unit 124, and a knowledge information storage unit. It has 125 and. The storage unit 120 stores various information, not limited to the above. The storage unit 120 may store information indicating a specific language corresponding to each language. The storage unit 120 is a semantic analyzer that outputs information on a semantic frame such as a specified domain goal and a score (also referred to as a “semantic analysis score”) indicating its accuracy (confidence) in response to input of character information. Remember. The storage unit 120 stores the information of the semantic analyzer for each specific language capable of semantic analysis. For example, the storage unit 120 stores information of a semantic analyzer for each specific language, such as an English semantic analyzer or a Japanese semantic analyzer, which is a specific language.

The language information storage unit 121 according to the embodiment stores various information related to the language. For example, the language information storage unit 121 stores various information in a language in which the information processing system 1 can identify the language (speech recognition). The language information storage unit 121 stores information indicating whether each language is a language capable of semantic analysis (specific language) and information indicating a language capable of translating each language (translation destination language). FIG. 4 is a diagram showing an example of the language information storage unit according to the embodiment. The language information storage unit 121 shown in FIG. 4 includes items such as "language" and "translation destination language". Further, the "translation destination language" includes items such as "# 1" and "# 2". Although only "# 1" and "# 2" are shown in FIG. 4, the "translation destination language" includes a number of items corresponding to the translation destination language such as "# 3" and "# 4". It may be.

"Language" indicates the language. For example, "language" indicates a language in which the information processing system 1 can identify the language (speech recognition). In addition, identification information (language ID) for identifying a language may be stored in association with each language. For example, a language code that identifies each language may be stored. "Translation destination language" indicates a language (translation destination language) capable of translating the language. For example, "translation destination language" indicates a translation destination language (translation destination language) in which the language can be translated.

In the example of FIG. 4, it is shown that the language "English" can be translated into Chinese, Hindi, etc. In addition, the language "Chinese" indicates that it can be translated into English, Arabic, or the like.

The language information storage unit 121 is not limited to the above, and may store various information depending on the purpose. For example, the language information storage unit 121 may store information indicating whether each language is a language (specific language) capable of semantic analysis. The language information storage unit 121 may store a flag indicating whether each language is a language (specific language) capable of semantic analysis. For example, the language information storage unit 121 may store the case where the language is a specific language in association with the flag "1" and the case where the language is not a specific language in association with the flag "0".

In this case, the information processing apparatus 100 may extract a specific language from the translation destination language associated with each language. For example, the information processing apparatus 100 may use a language whose flag is "1" among the translation destination languages associated with each language as a specific language.

The semantic frame information storage unit 122 according to the embodiment stores various information related to the semantic frame. The semantic frame information storage unit 122 stores various information related to the semantic frame for each language. The semantic frame information storage unit 122 stores information about the semantic frame corresponding to each language. For example, the semantic frame information storage unit 122 stores information about the semantic frame corresponding to each specific language.

In the example of FIG. 5, the semantic frame information storage unit 122 stores information (table) for each specific language, such as frame information FM1 and frame information FM2. For example, the frame information FM1 indicates information about a semantic frame of the language "English". Further, for example, the frame information FM2 indicates information regarding a semantic frame of the language "China".

The frame information FM1 and the frame information FM2 shown in FIG. 5 include items such as "language", "Domain-Goal", and "Slot". Further, "Slot" includes items such as "Attribute" and "Value".

"Language" indicates the language. For example, "language" indicates a language in which the information processing system 1 can identify the language (speech recognition). In addition, identification information (language ID) for identifying a language may be stored in association with each language. For example, a language code that identifies each language may be stored.

Also, "Domain-Goal" indicates the domain goal of the semantic frame. For example, "Domain-Goal" indicates the purpose (intention) of the utterance.

"Slot" stores various information about the corresponding Domain-Goal slot (component). For example, in "Slot", an attribute (slot name) included in the corresponding domain goal and its value (slot value) are stored. "Attribute" indicates an attribute (slot name) of a slot (component). "Value" indicates the slot value of the corresponding attribute (slot name). The “− (hyphen)” indicated by “Value” in the semantic frame information storage unit 122 indicates that the value is not stored in “Value”. When used in the processing of the user's semantic analysis, the "Value" stores a specific value (information) corresponding to the user's utterance.

In the example of FIG. 5, it is shown that the language "English" includes a meaning frame in which Domain-Goal is "Weather-Check" or "Music-Play". Further, it is shown that the Domain-Goal "Weather-Check" includes a slot whose "Attribute" is "Date" or "Place". That is, the Domain-Goal "Weather-Check" for checking the weather includes slots related to the date and time and place.

Note that the semantic frame information storage unit 122 is not limited to the above, and various information may be stored depending on the purpose. For example, the semantic frame information storage unit 122 may store the format of the value stored in each “Value”. For example, information indicating whether the value stored in "Value" is a numerical value or other information (character string or the like) may be stored. For example, information indicating whether the value stored in "Value" is information that can be commonly used in the language may be stored. For example, when the value stored in "Value" is a numerical value, information indicating that the information can be commonly used in the language may be stored.

The analysis accuracy information storage unit 123 according to the embodiment stores various information related to the analysis accuracy. The analysis accuracy information storage unit 123 stores various information related to the score, analysis accuracy, and the like in association with each character information. The analysis accuracy information storage unit 123 stores various information related to the score, analysis accuracy, and the like in association with each character information corresponding to the input language and the translation destination language. FIG. 6 is a diagram showing an example of the analysis accuracy information storage unit according to the embodiment. In the analysis accuracy information storage unit 123 shown in FIG. 6, "input language", "utterance development", "translation destination specific language", "translation destination translation sentence", "semantic analysis score", "semantic analysis accuracy (%)) ] Is included.

"Input language" indicates a language that can be input. For example, "input language" indicates a language in which the information processing system 1 can identify a language (speech recognition). In addition, identification information (language ID) for identifying the input language may be stored in association with each input language. For example, a language code that identifies each input language may be stored.

"Utterance development" indicates character information corresponding to the user's utterance and a paraphrase in which the character information is paraphrased. In the example of FIG. 6, among the "utterance development", "A" which is "play the Beatles" indicates the character information corresponding to the user's utterance, and the remaining three "B" to "D" are " Indicates a paraphrase that paraphrases "A".

"A", which is "Play the Beatles" in FIG. 6, may be described as "utterance development information A". For example, the utterance development information A indicates the character information of the user's utterance text as it is. Further, "B" which is "playing the Beatles song" in FIG. 6 may be described as "utterance development information B". For example, the utterance development information B clarifies the target as "song" in the utterance development information A, and indicates character information paraphrased in detail.

In addition, "C", which is "Please play the Beatles song" in Fig. 6, may be described as "Utterance development information C". "D", which is "Please play the Beatles song" in FIG. 6, may be described as "utterance development information D". For example, the utterance development information C and the utterance development information D indicate character information obtained by paraphrasing the utterance development information A into expressions used in more daily life. Note that FIG. 6 shows a case where the utterance development information A is paraphrased into three paraphrases of the utterance development information B, the utterance development information C, and the utterance development information D, but the number of paraphrases is four or more. It may be one or two.

"Translation destination specific language" indicates the translation destination language. For example, "translation destination specific language" is a language to be translated and indicates a specific language. For example, a language other than the specific language (non-specific language) is stored in the "translation destination specific language". In addition, identification information (language ID) for identifying the translation destination specific language may be stored in association with each translation destination specific language. For example, a language code that identifies each translation destination specific language may be stored.

In the example of FIG. 6, when the input language is also a specific language, the input language is also shown in a list together with the translation destination language in the "translation destination specific language". The same language as "indicates the target language, not the target language.

"Translation destination translation" indicates character information corresponding to the translation destination specific language. For example, the "translation destination translation sentence" indicates the character information in which the character information shown in the utterance development is translated into the corresponding translation destination language. For example, in the "translation destination translation sentence", the line corresponding to "utterance development" is utterance development information A and "translation destination specific language" corresponds to "English" is utterance development information A which is Japanese character information. Indicates the textual information translated into English.

In the example of FIG. 6, when the input language is also a specific language, the input language is also shown in a list together with the translation destination language in the "translation destination translation sentence". The character information corresponding to the same language as "" is not the translated sentence, but the same character information as the character-order information shown in "Utterance development".

"Semantic analysis score" indicates the degree of certainty (accuracy) of semantic analysis. The "semantic analysis score" indicates the certainty (accuracy) of the domain goal specified in the semantic analysis process.

"Semantic analysis accuracy (%)" indicates an accuracy index value that makes it possible to compare the accuracy of semantic analysis processing between multiple languages. The “semantic analysis accuracy (%)” indicates an index value (score) obtained by converting the value of the “semantic analysis score” by a predetermined function.

In the example of FIG. 6, the utterance development information A in which the input language "Japanese" is "playing the Beatles" corresponding to the user's utterance, and the utterance development information B to D in which the utterance development information A is paraphrased. The case where the utterance is expanded into three paraphrases is shown. In this way, the information processing device 100 performs utterance development that converts one character information corresponding to the user's utterance into a plurality of paraphrases.

In addition, the translation destination specific language of "Japanese", which is the input language (target language), is English, Spanish, French, German, Italian, Chinese (simplified), Korean, Hindi, Arabic, etc. Indicates that there is. In this way, the information processing apparatus 100 uses the character information of the target language "Japanese" in English, Spanish, French, German, Italian, Chinese (simplified), Korean, Hindi, Arabic, and the like. Convert to textual information.

In addition, the character information "Play the Beatles", which is the Japanese utterance development information A, indicates that the semantic analysis score is "0.83" and the semantic analysis accuracy is "82.9". In addition, the character information "Play the Beatles", which is the Japanese utterance development information A "Play the Beatles" translated into English, has a semantic analysis score of "0.51" and a semantic analysis accuracy of "0.51". It shows that it is 49.8 ”.

The analysis accuracy information storage unit 123 is not limited to the above, and may store various information depending on the purpose.

The threshold information storage unit 124 according to the embodiment stores various information related to the threshold value. The threshold information storage unit 124 stores various information regarding the threshold value used for comparison with the score. FIG. 7 is a diagram showing an example of the threshold information storage unit according to the embodiment. The threshold information storage unit 124 shown in FIG. 7 includes items such as “threshold ID” and “threshold”.

"Threshold ID" indicates identification information for identifying the threshold value. Further, the "threshold value" indicates a specific value of the threshold value identified by the corresponding threshold ID. In addition, information indicating its use is stored in association with each threshold value.

In the example of FIG. 7, it is shown that the value of the threshold value TH1 identified by the threshold value ID “TH1” is “0.75”. Further, the threshold value TH1 is stored in association with information indicating that its use is quality estimation (for example, translation).

The threshold information storage unit 124 is not limited to the above, and may store various information depending on the purpose. For example, the threshold information storage unit 124 may store the use of the threshold value in association with the threshold value ID. For example, the threshold information storage unit 124 may store the threshold ID “TH1” in association with the use “quality estimation”.

The knowledge information storage unit 125 according to the embodiment stores various information related to knowledge. The knowledge information storage unit 125 functions as a knowledge database (knowledge DB). The knowledge information storage unit 125 stores information in a multilingual dictionary. For example, the knowledge information storage unit 125 stores information indicating the correspondence between character strings indicating each object between languages. When the information processing device 100 acquires knowledge from the outside, the information processing device 100 does not have to have the knowledge information storage unit 125.

Return to Fig. 3 and continue the explanation. In the control unit 130, for example, a program stored inside the information processing apparatus 100 by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like (for example, a determination program such as an information processing program according to the present disclosure) is stored in a RAM. It is realized by executing (Random Access Memory) etc. as a work area. Further, the control unit 130 is a controller, and is realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 3, the control unit 130 includes an acquisition unit 131, a conversion unit 132, an execution unit 133, a calculation unit 134, a selection unit 135, an inverse conversion unit 136, a generation unit 137, and a transmission unit. It has 138 and realizes or executes the functions and actions of information processing described below. The internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 3, and may be another configuration as long as it is a configuration for performing information processing described later. Further, the connection relationship of each processing unit included in the control unit 130 is not limited to the connection relationship shown in FIG. 3, and may be another connection relationship.

The acquisition unit 131 acquires various information. The acquisition unit 131 acquires various information from an external information processing device. The acquisition unit 131 acquires various information from the terminal device 10.

The acquisition unit 131 acquires various information from the storage unit 120. The acquisition unit 131 acquires various information from the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125.

The acquisition unit 131 acquires various information converted by the conversion unit 132. The acquisition unit 131 acquires various information executed by the execution unit 133. The acquisition unit 131 acquires various information calculated by the calculation unit 134. The acquisition unit 131 acquires various information selected by the selection unit 135. The acquisition unit 131 acquires various information converted by the inverse conversion unit 136. The acquisition unit 131 acquires various information generated by the generation unit 137.

For example, the acquisition unit 131 may acquire a model (function). For example, the acquisition unit 131 acquires a model for estimating the quality of translation (quality estimation model). The acquisition unit 131 acquires a model (quality estimation model) for estimating the translation quality from an external information processing device or a storage unit 120 that provides various models (functions).

The acquisition unit 131 acquires the user's utterance information in the target language. The acquisition unit 131 acquires character information corresponding to the user's utterance in the target language.

The conversion unit 132 converts various information. The conversion unit 132 converts various information based on the information from the external information processing device and the information stored in the storage unit 120. The conversion unit 132 converts various information from the storage unit 120. The conversion unit 132 converts various information based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. ..

The conversion unit 132 converts the character information corresponding to the target language into the character information corresponding to the translation destination language to be the translation destination of the target language. The conversion unit 132 converts the character information corresponding to the target language into the character information corresponding to the language whose meaning can be interpreted.

The conversion unit 132 determines various information. The conversion unit 132 determines various information. The conversion unit 132 determines whether the language is capable of language identification (speech recognition). The conversion unit 132 determines whether the input language is a non-target language. The conversion unit 132 determines whether the target language is a non-target language. The conversion unit 132 determines that a language that cannot be language-identified (speech recognition) is a non-target language. The conversion unit 132 determines whether or not the input language is a specific language. The conversion unit 132 determines whether or not the target language is a specific language.

Execution unit 133 executes various processes. The execution unit 133 executes various processes based on information from an external information processing device. The execution unit 133 executes various processes based on the information stored in the storage unit 120. The execution unit 133 executes various processes based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. .. The execution unit 133 generates various information by executing the process.

The execution unit 133 executes various processes based on various information acquired by the acquisition unit 131. The execution unit 133 executes various processes based on various information converted by the conversion unit 132. The execution unit 133 executes various processes based on various information calculated by the calculation unit 134. The execution unit 133 executes various processes based on various information selected by the selection unit 135. The execution unit 133 executes various processes based on various information converted by the inverse conversion unit 136.

Execution unit 133 determines various information. The execution unit 133 determines various information. The execution unit 133 determines the execution of various processes. The execution unit 133 determines the execution of various processes. The execution unit 133 specifies various types of information. The execution unit 133 estimates various information. The execution unit 133 extracts various information. The execution unit 133 selects various information. The execution unit 133 extracts various information based on the information from the external information processing device and the information stored in the storage unit 120. The execution unit 133 extracts various information from the storage unit 120. The execution unit 133 extracts various information from the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125.

The execution unit 133 extracts various information based on the various information acquired by the acquisition unit 131. The execution unit 133 extracts various information based on the various information converted by the conversion unit 132. The execution unit 133 extracts various information based on the various information calculated by the calculation unit 134. The execution unit 133 extracts various information based on the various information selected by the selection unit 135. Further, the execution unit 133 extracts various information based on the various information converted by the inverse conversion unit 136. The execution unit 133 extracts various information based on the information generated by the generation unit 137.

Execution unit 133 executes analysis of character information corresponding to the user's utterance by appropriately using natural language processing technology such as morphological analysis. The execution unit 133 estimates (identifies) the content of the user's utterance by semantic analysis using the character information corresponding to the user's utterance. The execution unit 133 may estimate (specify) the user's situation by dialogue state estimation (Dialog State Tracking) using the character information corresponding to the user's utterance. The execution unit 133 estimates (identifies) the content of the character information by analyzing the character information converted by the conversion unit 132 by appropriately using semantic analysis and dialogue state estimation. The execution unit 133 analyzes the character information converted from the target language to the translation destination language by the conversion unit 132 by appropriately using a natural language processing technique such as morphological analysis. For example, the execution unit 133 estimates the content of the user's utterance corresponding to the character information by appropriately analyzing the character information by using various conventional techniques such as parsing.

The execution unit 133 estimates the content such as the intention of the user's utterance by analyzing the user's utterance. The execution unit 133 estimates the content such as the intention of the user's utterance by appropriately using various conventional techniques. The execution unit 133 estimates the content of the user's utterance by analyzing the user's utterance by appropriately using various conventional techniques. The execution unit 133 extracts important keywords from the character information of the user's utterance, and estimates the content of the user's utterance based on the extracted keywords.

The execution unit 133 identifies the Domain-Goal (domain goal) corresponding to the user's utterance by analyzing the character information corresponding to the utterance. The execution unit 133 estimates the attribute information such as the slot value corresponding to the specified Domain-Goal (domain goal) by analyzing the character information corresponding to the utterance. The execution unit 133 performs translation quality estimation (also simply referred to as “quality estimation”). Execution unit 133 estimates the quality of the translated character information. The execution unit 133 calculates the quality estimation translation accuracy (quality score) of the character information (translation text) by an appropriate method. For example, the execution unit 133 uses a model (quality estimation model) that outputs a score (quality score) indicating the quality estimation translation accuracy in response to the input of the character information before translation and the character information after translation. , Quality estimation translation accuracy (quality score) may be calculated. For example, the execution unit 133 uses a quality estimation model learned by using a combination of character information before translation, character information after translation, and its score (correct answer score) as learning data to determine the quality estimation translation accuracy (quality score). It may be calculated. The execution unit 133 uses a quality estimation model learned as learning data including a score (correct answer score) set by the administrator of the information processing system 1 based on the character information before translation and the character information after translation. It may be used to calculate the quality estimation translation accuracy (quality score). Execution unit 133 compares the quality score with the threshold value (for example, 0.75 etc.). If the quality score of the character information is equal to or higher than the threshold value, the execution unit 133 determines that the quality of the character information is high (high score), and if the quality estimation translation accuracy (quality score) of the translated text is less than the threshold value. If so, it is determined that the quality of the translated text is low (low score).

Execution unit 133 executes semantic analysis processing for one or more character information corresponding to each of one or more languages including the target language which is the language corresponding to the user's utterance. The execution unit 133 executes a semantic analysis process on one or more character information including one character information corresponding to the user's utterance in the target language. The execution unit 133 executes a semantic analysis process on one or more character information including the translated character information in which one character information is converted into the translation destination language to be the translation destination of the target language.

Execution unit 133 executes a semantic analysis process on one or more character information including a paraphrase in which one character information of the target language is paraphrased into another expression of the target language. The execution unit 133 executes a semantic analysis process on one or more character information including the translation paraphrase in which the paraphrase of the target language is converted into the translation destination language to which the target language is translated.

The execution unit 133 executes a semantic analysis process on the character information converted by the conversion unit 132. When the accuracy index value corresponding to the target language is equal to or higher than a predetermined value, the execution unit 133 does not execute the semantic analysis process of a language other than the target language in order to reduce the processing cost. When the accuracy index value corresponding to the target language is equal to or higher than a predetermined value, and when the target language is language identifiable, the execution unit 133 provides one or more character information corresponding to each of the one or more languages including the target language. On the other hand, the semantic analysis process is executed.

If the quality score indicating the translation quality of the character information is lower than the quality threshold value, the execution unit 133 does not execute the semantic analysis process using the character information. When the post-editing by a person is completed within a predetermined time, the execution unit 133 executes the semantic analysis process using the character information generated by the post-editing. When the post-editing by a person is not completed within a predetermined time, the execution unit 133 executes the process related to the process interruption. The execution unit 133 executes a process of notifying that the process is interrupted. When the language corresponding to the user's utterance can execute the semantic analysis process, the execution unit 133 executes the semantic analysis process on the character information of the target language spoken by the user.

Execution unit 133 executes the semantic analysis process for each language using the semantic analyzer of each language. The execution unit 133 uses a semantic analyzer that outputs information on a semantic frame such as a specified domain goal and a score (semantic analysis score) indicating its accuracy (confidence) in response to input of character information. Executes semantic analysis processing.

The execution unit 133 calculates a score (semantic analysis score) indicating the accuracy of the semantic analysis. The execution unit 133 calculates a semantic analysis score indicating the certainty (accuracy) of the domain goal specified in the semantic analysis process. The execution unit 133 may use the score output by the semantic analyzer used for the semantic analysis process as the semantic analysis score. The execution unit 133 uses the score output by the semantic analyzer for each language used for the semantic analysis processing of each language as the semantic analysis score of each language. The execution unit 133 may calculate the semantic analysis score by appropriately using various techniques.

The calculation unit 134 calculates various information. The calculation unit 134 calculates various values. The calculation unit 134 calculates various scores. For example, the calculation unit 134 calculates various types of information based on information from an external information processing device and information stored in the storage unit 120. The calculation unit 134 calculates various types of information based on information from other information processing devices such as the terminal device 10. The calculation unit 134 calculates various information based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. ..

The calculation unit 134 calculates various information based on various information acquired by the acquisition unit 131. The calculation unit 134 calculates various information based on the various information converted by the conversion unit 132. The calculation unit 134 calculates various information based on various processes executed by the execution unit 133. The calculation unit 134 calculates various information based on the result of the semantic analysis executed by the execution unit 133. The calculation unit 134 calculates various information based on the various information selected by the selection unit 135. The calculation unit 134 calculates various information based on the various information converted by the inverse conversion unit 136. The calculation unit 134 calculates various information based on various information generated by the generation unit 137.

The calculation unit 134 makes it possible to compare the accuracy of the semantic analysis processing for each of the one or more character information between a plurality of languages based on the result of the semantic analysis processing corresponding to each of the one or more character information. Calculate the index value. The calculation unit 134 calculates each accuracy index value of one or more character information by using a function that outputs the accuracy index value by inputting the information of the result of the semantic analysis process. The calculation unit 134 calculates each accuracy index value of one or more character information by using a function that outputs an accuracy index value by inputting a score included in the result of the semantic analysis process. The calculation unit 134 uses a function that inputs a score (semantic analysis score) included in the result of the semantic analysis process and outputs an accuracy index value (semantic analysis accuracy) to obtain each accuracy index value of one or more character information. calculate. The calculation unit 134 uses a function that outputs the accuracy index value by inputting the score of the semantic analysis process for the character information of one language and the information indicating one language, and the accuracy index value of the character information of one language. Is calculated. The calculation unit 134 calculates the accuracy index value corresponding to the character information of the target language.

The selection unit 135 selects various information. The selection unit 135 extracts various information. The selection unit 135 specifies various types of information. The selection unit 135 selects various information based on the information from the external information processing device and the information stored in the storage unit 120. The selection unit 135 selects various types of information based on information from other information processing devices such as the terminal device 10. The selection unit 135 selects various information based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. ..

The selection unit 135 selects various information based on various information acquired by the acquisition unit 131. The selection unit 135 selects various information based on the various information converted by the conversion unit 132. The selection unit 135 selects various information based on various processes executed by the execution unit 133. The selection unit 135 selects various information based on the result of the semantic analysis executed by the execution unit 133. The selection unit 135 selects various information based on the various information calculated by the calculation unit 134. The selection unit 135 selects various information based on the various information converted by the inverse conversion unit 136. The selection unit 135 selects various information based on the various information generated by the generation unit 137.

The selection unit 135 selects the processing target character information, which is the character information used for processing, from the one or more character information based on each accuracy index value of the one or more character information calculated by the calculation unit 134. The selection unit 135 selects the character information having the maximum accuracy index value as the character information to be processed. When the accuracy index value corresponding to the character information of the target language is equal to or higher than a predetermined value, the selection unit 135 selects the character information of the target language as the character information to be processed.

The inverse conversion unit 136 converts various information. For example, the inverse conversion unit 136 converts various information based on the information from the external information processing device and the information stored in the storage unit 120. The inverse conversion unit 136 converts various information based on the information from other information processing devices such as the terminal device 10. The inverse conversion unit 136 converts various information based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. do.

The inverse conversion unit 136 converts various information based on various information acquired by the acquisition unit 131. The inverse conversion unit 136 converts various information based on the various information converted by the conversion unit 132. The inverse conversion unit 136 converts various information based on the various information calculated by the execution unit 133. The inverse conversion unit 136 converts various information based on the various information calculated by the calculation unit 134. The inverse conversion unit 136 converts various information based on the various information selected by the selection unit 135. The inverse conversion unit 136 converts various information based on the various information generated by the generation unit 137. The inverse conversion unit 136 changes various information based on the conversion. Various information is updated based on the information acquired by the acquisition unit 131.

The inverse transformation unit 136 converts the result of the semantic analysis processing corresponding to the language of the processing target character information selected by the selection unit 135 into the target language. The inverse transformation unit 136 converts the result of the semantic analysis processing corresponding to the language of the processing target character information into the target language. When the language of the character information to be processed is other than the target language, the inverse transformation unit 136 converts the result of the semantic analysis process into the target language. The inverse transformation unit 136 converts a part of the result of the semantic analysis process into the target language. The inverse conversion unit 136 converts the slot value of the result of the semantic analysis process into the target language.

The inverse conversion unit 136 determines various information. The inverse conversion unit 136 determines various information. The inverse conversion unit 136 determines the execution of various processes. The inverse transformation unit 136 determines whether or not the inverse transformation needs to be executed. The inverse transformation unit 136 does not have to convert the information that does not require inverse transformation into the target language. The inverse transformation unit 136 does not have to convert the information that does not require inverse transformation among the results of the semantic analysis process into the target language. The inverse transformation unit 136 does not have to convert the information common to the languages among the results of the semantic analysis processing into the target language. The inverse conversion unit 136 does not have to convert the information common to the languages such as numerical values among the slot values into the target language.

Generation unit 137 generates various information. The generation unit 137 generates various information based on the information from the external information processing device and the information stored in the storage unit 120. The generation unit 137 generates various information based on the information from other information processing devices such as the terminal device 10. The generation unit 137 generates various information based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. ..

The generation unit 137 generates various information based on the various information acquired by the acquisition unit 131. The generation unit 137 generates various information based on the various information converted by the conversion unit 132. The generation unit 137 generates various information based on various information generated by the processing execution of the execution unit 133. The generation unit 137 generates various information based on the various information calculated by the calculation unit 134. The generation unit 137 generates various information based on the various information selected by the selection unit 135. The generation unit 137 generates various information based on the various information converted by the inverse conversion unit 136.

The generation unit 137 appropriately uses various techniques to generate various information such as a screen (image information) to be provided to an external information processing device. The generation unit 137 generates a screen (image information) or the like to be provided to the terminal device 10. For example, the generation unit 137 generates a screen (image information) or the like to be provided to the terminal device 10 based on the information stored in the storage unit 120.

The generation unit 137 may generate the screen (image information) or the like by any process as long as the screen (image information) or the like to be provided to the external information processing device can be generated. For example, the generation unit 137 generates a screen (image information) to be provided to the terminal device 10 by appropriately using various techniques related to image generation, image processing, and the like. For example, the generation unit 137 appropriately uses various techniques such as Java (registered trademark) to generate a screen (image information) to be provided to the terminal device 10. The generation unit 137 may generate a screen (image information) to be provided to the terminal device 10 based on the format of CSS, Javascript (registered trademark), or HTML. Further, for example, the generation unit 137 may generate a screen (image information) in various formats such as JPEG (Joint Photographic Experts Group), GIF (Graphics Interchange Format), and PNG (Portable Network Graphics).

For example, the generation unit 137 generates a quality estimation model using learning data including a combination of character information before translation, character information after translation, and a score (correct answer score) thereof. The generation unit 137 uses learning data including a score (correct answer score) set by the administrator of the information processing system 1 based on the character information before translation and the character information after translation to generate a quality estimation model. Generate.

The transmission unit 138 transmits various information. The transmission unit 138 transmits various information to an external information processing device. The transmission unit 138 provides various information to an external information processing device. For example, the transmission unit 138 transmits various information to another information processing device such as the terminal device 10. The transmission unit 138 provides the information stored in the storage unit 120. The transmission unit 138 transmits the information stored in the storage unit 120.

The transmission unit 138 provides various types of information based on information from other information processing devices such as the terminal device 10. The transmission unit 138 provides various information based on the information stored in the storage unit 120. The transmission unit 138 provides various information based on the information stored in the language information storage unit 121, the semantic frame information storage unit 122, the analysis accuracy information storage unit 123, the threshold information storage unit 124, and the knowledge information storage unit 125. ..

The transmission unit 138 transmits information indicating a function to be executed by the terminal device 10 to the terminal device 10. The transmission unit 138 transmits information indicating the function (service) selected by the selection unit 135 to the terminal device 10. The transmission unit 138 transmits various information to the terminal device 10 in response to an instruction from the execution unit 133. The transmission unit 138 transmits information instructing the terminal device 10 to execute a function (service). The transmission unit 138 transmits the image information generated by the generation unit 137.

[1-4. Configuration of terminal device according to embodiment]
Next, the configuration of the terminal device 10 which is an example of the information processing device that executes the information processing according to the embodiment will be described. FIG. 8 is a diagram showing a configuration example of the terminal device according to the embodiment of the present disclosure.

As shown in FIG. 8, the terminal device 10 includes a communication unit 11, an input unit 12, an output unit 13, a storage unit 14, a control unit 15, a sensor unit 16, and a display unit 17.

The communication unit 11 is realized by, for example, a NIC or a communication circuit. The communication unit 11 is connected to the network N (Internet or the like) by wire or wirelessly, and transmits / receives information to / from other devices such as the information processing device 100 via the network N.

The input unit 12 accepts various inputs. The input unit 12 receives the detection by the sensor unit 16 as an input. The input unit 12 accepts the input of the user's utterance information. The input unit 12 accepts input by the user's physical movement. The input unit 12 accepts the user's gesture and line of sight as input.

Various operations are input from the user to the input unit 12. The input unit 12 receives sound as input by the sensor unit 16 having a function of detecting voice. The input unit 12 receives the voice information detected by the microphone (sound sensor) that detects the voice as the input information. The input unit 12 receives the voice spoken by the user as input information.

Further, the input unit 12 may accept an operation (user operation) on the terminal device 10 used by the user as an operation input by the user. The input unit 12 may receive information regarding the operation of the user using the remote controller (remote controller) via the communication unit 11. Further, the input unit 12 may have a button provided on the terminal device 10 or a keyboard or mouse connected to the terminal device 10.

For example, the input unit 12 may have a touch panel capable of realizing functions equivalent to those of a remote controller, a keyboard, and a mouse. In this case, various information is input to the input unit 12 via the display unit 17. The input unit 12 receives various operations from the user via the display screen by the function of the touch panel realized by various sensors. That is, the input unit 12 receives various operations from the user via the display unit 17 of the terminal device 10. For example, the input unit 12 receives an operation such as a user's designated operation via the display unit 17 of the terminal device 10. For example, the input unit 12 functions as a reception unit that receives a user's operation by the function of the touch panel. In this case, the input unit 12 and the reception unit 153 may be integrated. As the detection method of the user's operation by the input unit 12, the capacitance method is mainly adopted in the tablet terminal, but other detection methods such as the resistance film method, the surface acoustic wave method, the infrared method, and the electromagnetic induction method are used. Any method may be adopted as long as the user's operation can be detected and the touch panel function can be realized.

For example, the input unit 12 accepts a user's utterance as an input. The input unit 12 receives the user's utterance detected by the sensor unit 16 as input. The input unit 12 receives the user's utterance detected by the sound sensor of the sensor unit 16 as an input.

The output unit 13 outputs various information. The output unit 13 has a function of outputting audio. For example, the output unit 13 has a speaker that outputs sound. The output unit 13 outputs various information by voice according to the control by the execution unit 152. The output unit 13 outputs information by voice to the user. The output unit 13 outputs the information displayed on the display unit 17 by voice.

The storage unit 14 is realized by, for example, a semiconductor memory element such as a RAM or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 14 stores various information used for displaying the information.

Return to Fig. 8 and continue the explanation. The control unit 15 is realized by, for example, a CPU, an MPU, or the like executing a program stored inside the terminal device 10 (for example, a display program such as an information processing program according to the present disclosure) with a RAM or the like as a work area. Will be done. Further, the control unit 15 is a controller, and may be realized by an integrated circuit such as an ASIC or FPGA.

As shown in FIG. 8, the control unit 15 includes a reception unit 151, an execution unit 152, a reception unit 153, and a transmission unit 154, and realizes or executes the information processing functions and operations described below. .. The internal configuration of the control unit 15 is not limited to the configuration shown in FIG. 8, and may be another configuration as long as it is a configuration for performing information processing described later.

The receiving unit 151 receives various information. The receiving unit 151 receives various information from an external information processing device. The receiving unit 151 receives various information from other information processing devices such as the information processing device 100.

The receiving unit 151 receives information instructing the execution of a function (service) from the information processing device 100. The receiving unit 151 receives execution instructions of various functions (services) from the information processing device 100. For example, the receiving unit 151 receives information specifying a function (service) from the information processing device 100 as a function execution instruction. The receiving unit 151 receives the content. The receiving unit 151 receives the content to be displayed from the information processing device 100.

Execution unit 152 executes various processes. The execution unit 152 determines the execution of various processes. The execution unit 152 executes various processes based on information from an external information processing device. The execution unit 152 executes various processes based on the information from the information processing device 100. The execution unit 152 executes various processes in response to an instruction from the information processing device 100. The execution unit 152 executes various processes based on the information stored in the storage unit 14. The execution unit 152 executes a function (service).

The execution unit 152 controls various outputs. The execution unit 152 controls the audio output by the output unit 13. The execution unit 152 controls various displays. The execution unit 152 controls the display of the display unit 17. The execution unit 152 controls the display of the display unit 17 in response to the reception by the reception unit 151. The execution unit 152 controls the display of the display unit 17 based on the information received by the reception unit 151. The execution unit 152 controls the display of the display unit 17 based on the information received by the reception unit 153. The execution unit 152 controls the display of the display unit 17 in response to the reception by the reception unit 153.

Reception department 153 receives various information. The reception unit 153 receives input by the user via the input unit 12. The reception unit 153 accepts the utterance by the user as an input. The reception unit 153 accepts operations by the user. The reception unit 153 accepts the user's operation on the information displayed by the display unit 17. The reception unit 153 accepts character input by the user.

The transmission unit 154 transmits various information to an external information processing device. For example, the transmission unit 154 transmits various information to another information processing device such as the information processing device 100. The transmission unit 154 transmits the information stored in the storage unit 14.

The transmission unit 154 transmits various types of information based on information from other information processing devices such as the information processing device 100. The transmission unit 154 transmits various types of information based on the information stored in the storage unit 14.

The transmission unit 154 transmits the sensor information detected by the sensor unit 16 to the information processing device 100. The transmission unit 154 transmits the user's utterance information detected by the sound sensor of the sensor unit 16 to the information processing device 100.

The transmission unit 154 transmits the input information input by the user to the information processing device 100. The transmission unit 154 transmits the input information voice-input by the user to the information processing device 100. The transmission unit 154 transmits the input information input by the user's operation to the information processing device 100. The transmission unit 154 transmits the user's utterance information in the target language to the information processing device 100. The transmission unit 154 transmits the character information corresponding to the user's utterance in the target language to the information processing device 100.

The sensor unit 16 detects various sensor information. The sensor unit 16 has a sound sensor (speaker) that detects sound. The sensor unit 16 has a function as an imaging unit for capturing an image. The sensor unit 16 has an image sensor function and detects image information. The sensor unit 16 functions as an image input unit that receives an image as an input. The sensor unit 16 is not limited to the above, and may have various sensors. The sensor unit 16 is a position sensor, an acceleration sensor, a gyro sensor, a temperature sensor, a humidity sensor, an illuminance sensor, a pressure sensor, a proximity sensor, a sensor for receiving biological information such as odor, sweat, heartbeat, pulse, and brain wave. It may have various sensors. Further, the sensors that detect the above-mentioned various information in the sensor unit 16 may be common sensors, or may be realized by different sensors.

The display unit 17 is provided on the terminal device 10 and displays various information. The display unit 17 is realized by, for example, a liquid crystal display, an organic EL (Electro-Luminescence) display, or the like. The display unit 17 may be realized by any means as long as the information provided by the information processing device 100 can be displayed. The display unit 17 displays various information according to the control by the execution unit 152.

The display unit 17 displays various information received by the reception unit 151. The display unit 17 displays the response received from the information processing device 100. The display unit 17 displays information related to language conversion.

[1-5. Response example]
Here, a response example will be described with reference to FIG. FIG. 9 is a diagram showing an example of the response according to the embodiment of the present disclosure. FIG. 9 shows an example of a response in a language-recognizable format.

The information processing system 1 may output information for the user to recognize when the input language and the language for which the semantic analysis processing is performed are different. For example, the information processing system 1 translates the input language and performs semantic analysis processing using the translated information in three phases of input, analysis, and output at the time of response. Output in a format that shows which language was processed.

For example, the terminal device 10 displays a response in a language-recognizable format on the display unit 17. In the example of FIG. 9, in the information processing system 1, the input language (target language) is Burmese, the specific language (translation destination language) for which semantic analysis is performed is Japanese, and the output language (target language) is Myanmar. Outputs information indicating that it is a word. In this way, the information processing system 1 outputs a response in a format that indicates in which language the processing was performed in three phases of input, analysis, and output.

[1-6. Information processing procedure according to the embodiment]
Next, various information processing procedures according to the embodiment will be described with reference to FIGS. 10 to 12.

[1-6-1. Procedure for processing related to information processing equipment]
First, the flow of processing related to the information processing apparatus according to the embodiment of the present disclosure will be described with reference to FIG. FIG. 10 is a flowchart showing processing of the information processing apparatus according to the embodiment of the present disclosure. Specifically, FIG. 10 is a flowchart showing a procedure of information processing by the information processing apparatus 100.

As shown in FIG. 10, the information processing apparatus 100 executes a semantic analysis process on the character information corresponding to each of the languages including the target language corresponding to the user's utterance (step S101). Then, the information processing apparatus 100 calculates an accuracy index value for each of the character information based on the result of the semantic analysis process (step S102). For example, the information processing apparatus 100 calculates an accuracy index value (semantic analysis accuracy) that makes it possible to compare the accuracy of the semantic analysis process between a plurality of languages for each of the character information.

[1-6-2. Procedure for processing related to information processing system]
Next, an example of specific processing related to the information processing system will be described with reference to FIG. FIG. 11 is a flowchart showing processing of the information processing system according to the embodiment of the present disclosure. In the following, a case where the information processing system 1 performs processing will be described as an example, but the processing shown in FIG. 11 is performed by any of the information processing device 100 and the terminal device 10 included in the information processing system 1. May be good.

As shown in FIG. 11, the information processing system 1 acquires the voice information of the utterance by the user (step S201). For example, the information processing system 1 acquires voice information of a user's utterance in an input language (target language).

Then, the information processing system 1 performs voice recognition processing (step S202). The information processing system 1 performs voice recognition processing for voice information spoken by the user. For example, the information processing system 1 acquires the utterance text (character information) of the utterance by the user in the input language (target language) as the utterance information by voice recognition. For example, the information processing system 1 determines whether the language is capable of language identification (speech recognition), and if possible, performs voice recognition processing. If the information processing system 1 is not a language capable of language identification (speech recognition), the processing may be terminated. In this case, the information processing system 1 may notify the user that the language is not compatible. For example, the information processing system 1 uses the utterance information as input language character information. When it is necessary to normalize the utterance information, the information processing system 1 may normalize the utterance information to the input language character information.

Then, the information processing system 1 develops the utterance sentence (step S203). The information processing system 1 generates a paraphrase in which the input language character information corresponding to the user's utterance is paraphrased. The information processing system 1 generates a pseudo-utterance text list including input language character information corresponding to the user's utterance and its paraphrase. For example, as shown in FIG. 6, the information processing system 1 has three (plurality) parameters of utterance development information B, utterance development information C, and utterance development information D, which are paraphrases of utterance development information A corresponding to the user's utterance. Generate a phrase. As a result, the information processing system 1 generates a pseudo-utterance text list including input language character information corresponding to the user's utterance and its paraphrase.

Then, the information processing system 1 develops the language (step S204). The information processing system 1 translates (converts) the input language character information and its paraphrase into another language. The information processing system 1 translates the input language character information of the target language into another language. The information processing system 1 converts the input language character information into the character information of the translation destination language (translation destination character information). For example, as shown in FIG. 6, the information processing system 1 translates the utterance development information A, which is Japanese input language character information, into English, Spanish, French, and the like. Further, the information processing system 1 translates each paraphrase of the target language into another language. The information processing system 1 converts each paraphrase into character information (translation destination character information) of the translation destination language. For example, the information processing system 1 translates the utterance development information B, which is a Japanese paraphrase, into English, Spanish, French, and the like as shown in FIG. Similarly, as shown in FIG. 6, the information processing system 1 translates the utterance development information C and the utterance development information D, which are Japanese paraphrases, into English, Spanish, French, and the like. As a result, the information processing system 1 provides a list of translated sentences including the input language character information, its paraphrase, the translated sentence obtained by translating the input language character information into each language, and the translated sentence (translated paraphrase) obtained by translating the paraphrase. Generate.

Then, the information processing system 1 performs loop processing of the translated sentence list (step S205). The information processing system 1 selects each sentence (translated sentence, etc.) in the translated sentence list one by one, and processes the selected sentence (also referred to as "selected sentence").

The information processing system 1 performs utterance semantic analysis processing on the selected sentence (step S206). The information processing system 1 generates the result of the semantic analysis process for the selected sentence. For example, the information processing system 1 estimates information such as a domain goal of a selected sentence by a speech semantic analysis process, and generates information of a semantic frame including information such as a semantic analysis score indicating the certainty of the estimated information. ..

Then, the information processing system 1 performs analysis accuracy conversion based on the result of the utterance semantic analysis process and the score function (step S207). For example, the information processing system 1 performs analysis accuracy conversion using the equation (1). The information processing system 1 calculates the accuracy index value (semantic analysis accuracy) indicating the semantic analysis accuracy of the selected sentence by inputting the semantic analysis score and the information indicating the language of the selected sentence into the equation (1). Further, the information processing system 1 makes it possible to identify the sentence in the translated sentence list as a processed sentence by adding a flag to the sentence for which the calculation of the accuracy index value (semantic analysis accuracy) has been completed. The information processing system 1 may exclude sentences for which the accuracy index value (semantic analysis accuracy) has been calculated from the translated sentence list.

The information processing system 1 performs the processes of steps S206 to S207 with each sentence in the translated sentence list as a selected sentence by the loop processing of the translated sentence list. As a result, the information processing system 1 calculates the accuracy index value (semantic analysis accuracy) of each sentence in the translated sentence list by the loop processing of the translated sentence list.

When there is an unprocessed sentence in the translated sentence list (step S208: No), the information processing system 1 selects one unprocessed sentence in the translated sentence list and performs the processes of steps S206 to S207.

On the other hand, the information processing system 1 ends the loop processing of the translated sentence list when there is no unprocessed sentence in the translated sentence list (step S208: Yes).

Then, the information processing system 1 selects a semantic frame (step S209). The information processing system 1 selects the sentence having the maximum accuracy index value (semantic analysis accuracy) from each sentence as the sentence to be used in the subsequent processing. The information processing system 1 selects the semantic analysis result of the sentence having the maximum accuracy index value (semantic analysis accuracy) among the sentences as the sentence to be used in the subsequent processing. The information processing system 1 selects, among the sentences, the semantic frame corresponding to the sentence having the maximum accuracy index value (semantic analysis accuracy) as the sentence to be used in the subsequent processing.

Then, the information processing system 1 performs slot inverse transformation (step S210). The information processing system 1 performs slot inverse conversion using a sentence meaning frame (semantic frame with maximum accuracy) having the maximum accuracy index value (semantic analysis accuracy). The information processing system 1 converts the slot value in the meaning frame having the maximum accuracy into the slot value of the input language (target language). The information processing system 1 converts the slot value of the specific language (translation destination language) into the slot value of the input language (target language). The information processing system 1 does not need to perform step S210 when the inverse transformation is unnecessary, such as when the language of the sentence having the maximum accuracy index value (semantic analysis accuracy) is the target language.

Then, the information processing system 1 generates a response (step S211). The information processing system 1 generates information such as images, sounds, and texts according to the output mode. Then, the information processing system 1 outputs the generated information (step S212). The information processing system 1 displays images and texts and outputs sounds.

[1-6-3. Other Procedures for Processing Related to Information Processing Systems Part 1]
Next, another example of specific processing related to the information processing system will be described with reference to FIG. FIG. 12 is a flowchart showing processing of the information processing system according to the embodiment of the present disclosure. In the following, a case where the information processing system 1 performs processing will be described as an example, but the processing shown in FIG. 12 is performed by any of the information processing device 100 and the terminal device 10 included in the information processing system 1. May be good. The same points as in FIG. 11 will be omitted as appropriate.

As shown in FIG. 12, the information processing system 1 acquires the voice information of the utterance by the user (step S301). For example, the information processing system 1 acquires voice information of a user's utterance in an input language (target language).

Then, the information processing system 1 performs voice recognition processing (step S302). The information processing system 1 performs voice recognition processing for voice information spoken by the user. For example, the information processing system 1 acquires the utterance text (character information) of the utterance by the user in the input language (target language) as the utterance information by voice recognition. For example, the information processing system 1 determines whether the language is capable of language identification (speech recognition), and if possible, performs voice recognition processing. If the information processing system 1 is not a language capable of language identification (speech recognition), the processing may be terminated. In this case, the information processing system 1 may notify the user that the language is not compatible. For example, the information processing system 1 uses the utterance information as input language character information. When it is necessary to normalize the utterance information, the information processing system 1 may normalize the utterance information to the input language character information.

The information processing system 1 performs an utterance semantic analysis process on the input language character information for the utterance information (step S303). The information processing system 1 generates the result of the semantic analysis process for the input language character information. For example, the information processing system 1 estimates information such as a domain goal of input language character information by utterance meaning analysis processing, and provides information on a meaning frame including information such as a meaning analysis score indicating the certainty of the estimated information. Generate.

Then, when the accuracy of the semantic analysis process using the input language character information is low (step S304: Yes), the information processing system 1 develops the utterance sentence (step S305). For example, the information processing system 1 determines whether or not the accuracy of the semantic analysis process is low based on the comparison between the semantic analysis score of the semantic analysis process using the input language character information and a predetermined threshold value. When the semantic analysis score of the semantic analysis process using the input language character information is equal to or less than a predetermined threshold value, the information processing system 1 determines that the accuracy of the semantic analysis process is low. The information processing system 1 may determine whether or not the accuracy of the semantic analysis process is low by using the accuracy index value (semantic analysis accuracy) obtained by converting the semantic analysis score. In this case, the information processing system 1 determines that the accuracy of the semantic analysis process is low when the accuracy index value (semantic analysis accuracy) of the input language character information calculated by the equation (1) is equal to or less than a predetermined threshold value.

In step S305, the information processing system 1 generates a paraphrase in which the input language character information corresponding to the user's utterance is paraphrased. Since this point is the same as the utterance sentence development in FIG. 11, the description thereof will be omitted. The information processing system 1 generates a pseudo-utterance text list including input language character information corresponding to the user's utterance and its paraphrase.

Then, the information processing system 1 develops the language (step S306). The information processing system 1 translates (converts) the input language character information and its paraphrase into another language. In the example of FIG. 12, the information processing system 1 translates (converts) the input language character information and its paraphrase into another language by using the information of the specific language. The information processing system 1 uses information in a specific language to translate (convert) input language character information and its paraphrases into another language capable of semantic analysis processing. Since this point is the same as the language development in FIG. 11, the description thereof will be omitted. The information processing system 1 generates a translation sentence list including input language character information, its paraphrase, a translation sentence obtained by translating the input language character information into each language, and a translation sentence (translation paraphrase) obtained by translating the paraphrase.

Then, the information processing system 1 performs loop processing of the translated sentence list (step S307). The information processing system 1 selects each sentence (translated sentence, etc.) in the translated sentence list one by one, and processes the selected sentence (selected sentence).

The information processing system 1 performs utterance semantic analysis processing on the selected sentence (step S308). The information processing system 1 generates the result of the semantic analysis process for the selected sentence. For example, the information processing system 1 estimates information such as a domain goal of a selected sentence by a speech semantic analysis process, and generates information of a semantic frame including information such as a semantic analysis score indicating the certainty of the estimated information. ..

Then, the information processing system 1 performs analysis accuracy conversion based on the result of the utterance semantic analysis process and the score function (step S309). For example, the information processing system 1 performs analysis accuracy conversion using the equation (1). The information processing system 1 calculates the accuracy index value (semantic analysis accuracy) indicating the semantic analysis accuracy of the selected sentence by inputting the semantic analysis score and the information indicating the language of the selected sentence into the equation (1). Further, the information processing system 1 makes it possible to identify the sentence in the translated sentence list as a processed sentence by adding a flag to the sentence for which the calculation of the accuracy index value (semantic analysis accuracy) has been completed. The information processing system 1 may exclude sentences for which the accuracy index value (semantic analysis accuracy) has been calculated from the translated sentence list.

The information processing system 1 performs the processes of steps S308 to S309 with each sentence in the translated sentence list as a selected sentence by the loop processing of the translated sentence list. As a result, the information processing system 1 calculates the accuracy index value (semantic analysis accuracy) of each sentence in the translated sentence list by the loop processing of the translated sentence list.

When there is an unprocessed sentence in the translated sentence list (step S310: No), the information processing system 1 selects one unprocessed sentence in the translated sentence list and performs the processes of steps S308 to S309.

On the other hand, when there is no unprocessed sentence in the translated sentence list (step S310: Yes), the information processing system 1 ends the loop processing of the translated sentence list.

Then, the information processing system 1 selects a semantic frame (step S311). The information processing system 1 selects the sentence having the maximum accuracy index value (semantic analysis accuracy) from each sentence as the sentence to be used in the subsequent processing. The information processing system 1 selects the semantic analysis result of the sentence having the maximum accuracy index value (semantic analysis accuracy) among the sentences as the sentence to be used in the subsequent processing. The information processing system 1 selects, among the sentences, the semantic frame corresponding to the sentence having the maximum accuracy index value (semantic analysis accuracy) as the sentence to be used in the subsequent processing.

Then, the information processing system 1 performs slot inverse transformation (step S312). The information processing system 1 performs slot inverse conversion using a sentence meaning frame (semantic frame with maximum accuracy) having the maximum accuracy index value (semantic analysis accuracy). The information processing system 1 converts the slot value in the meaning frame having the maximum accuracy into the slot value of the input language (target language). The information processing system 1 converts the slot value of the specific language (translation destination language) into the slot value of the input language (target language). Note that the information processing system 1 does not have to perform step S312 when the inverse transformation is unnecessary, such as when the language of the sentence having the maximum accuracy index value (semantic analysis accuracy) is the target language.

Then, the information processing system 1 generates a response (step S313). The information processing system 1 generates information such as images, sounds, and texts according to the output mode. For example, the information processing system 1 generates a response using the result of the semantic analysis process of the sentence having the maximum accuracy index value (semantic analysis accuracy) of each sentence.

If the accuracy of the semantic analysis process using the input language character information is not low (step S304: No), the information processing system 1 generates a response using the result of the semantic analysis process of the input language character information. For example, when the accuracy of the semantic analysis process using the input language character information is high, the information processing system 1 responds by using the result of the semantic analysis process of the input language character information without performing the processes of steps S305 to S312. Generate.

Then, the information processing system 1 outputs the generated information (step S314). The information processing system 1 displays images and texts and outputs sounds.

[1-6-4. Other Procedures for Processing Related to Information Processing Systems Part 2]
Next, another example of specific processing related to the information processing system will be described with reference to FIG. FIG. 13 is a flowchart showing processing of the information processing system according to the embodiment of the present disclosure. In the following, a case where the information processing system 1 performs processing will be described as an example, but the processing shown in FIG. 13 is performed by any of the information processing device 100 and the terminal device 10 included in the information processing system 1. May be good. The same points as those in FIGS. 11 and 12 will be omitted as appropriate.

As shown in FIG. 13, the information processing system 1 acquires the voice information of the utterance by the user (step S401). For example, the information processing system 1 acquires voice information of a user's utterance in an input language (target language).

Then, the information processing system 1 performs voice recognition processing (step S402). The information processing system 1 performs voice recognition processing for voice information spoken by the user. For example, the information processing system 1 acquires the text (utterance information) of the utterance by the user in the input language (target language) as the utterance information by voice recognition.

Then, the information processing system 1 determines whether the language is not the target language (step S403). For example, the information processing system 1 determines that a language that cannot be language-identified (speech recognition) is a non-target language.

When the information processing system 1 determines that the language is not a non-target language (step S403: Yes), the information processing system 1 expands the utterance sentence (step S404). The information processing system 1 generates a paraphrase in which the input language character information corresponding to the user's utterance is paraphrased. Since this point is the same as the utterance sentence development in FIG. 11, the description thereof will be omitted. The information processing system 1 generates a pseudo-utterance text list including input language character information corresponding to the user's utterance and its paraphrase.

Then, the information processing system 1 develops the language (step S405). The information processing system 1 translates (converts) the input language character information and its paraphrase into another language. In the example of FIG. 12, the information processing system 1 translates (converts) the input language character information and its paraphrase into another language by using the information of the specific language. The information processing system 1 uses information in a specific language to translate (convert) input language character information and its paraphrases into another language capable of semantic analysis processing. Since this point is the same as the language development in FIG. 11, the description thereof will be omitted. The information processing system 1 generates a translation sentence list including input language character information, its paraphrase, a translation sentence obtained by translating the input language character information into each language, and a translation sentence (translation paraphrase) obtained by translating the paraphrase.

Then, the information processing system 1 performs loop processing of the translated sentence list (step S406). The information processing system 1 selects each sentence (translated sentence, etc.) in the translated sentence list one by one, and processes the selected sentence (selected sentence).

The information processing system 1 estimates the quality of the selected sentence (step S407). For example, the information processing system 1 estimates the quality of the selected sentence (translated text) when the selected sentence is a translated sentence. For example, the information processing system 1 calculates the quality estimated translation accuracy (quality score) of the translated text (translated text) by an appropriate method, and sets the quality estimated translation accuracy (quality score) and the threshold value (for example, 0.75, etc.). compare. Then, if the quality estimated translation accuracy (quality score) of the translated sentence is equal to or higher than the threshold value, the information processing system 1 determines that the quality of the selected sentence is high (high score), and determines that the quality estimated translation accuracy of the selected sentence is high. If (quality score) is less than the threshold value, it is determined that the quality of the selected sentence is low (low score).

When the information processing system 1 estimates (determines) that the quality is low (step S407: LOW), it determines whether manual editing of the selected sentence is completed within a predetermined time (step S408).

When the information processing system 1 determines that the manual editing is not completed within a predetermined time (step S408: LONG TIME), the information processing system 1 rejects the processing of the selected sentence (step S415). For example, when the information processing system 1 determines that the manual editing of the selected sentence is not completed within a predetermined time, the processing of the selected sentence is interrupted. Then, the information processing system 1 generates a response in step S415 by using the reason for interruption for the selected sentence. The information processing system 1 continues the loop processing of the translated sentence list in step S406 for the remaining sentences (unprocessed sentences) even when the processing of a certain selected sentence is rejected.

When the information processing system 1 determines that the manual editing of the selected sentence is completed within a predetermined time (step S408: SHORT TIME), the selected sentence is used in the manually edited character information in step S409. ~ S410 is processed. When the information processing system 1 determines that the manual editing of the selected sentence is completed within a predetermined time, the information processing system 1 replaces the selected sentence with the manually edited sentence (edited sentence).

Further, when the information processing system 1 estimates (determines) that the quality is high (step S407: HIGH), the information processing system 1 performs the processes of steps S409 to S410 using the selection statement.

The information processing system 1 performs utterance semantic analysis processing on the selected sentence (step S409). The information processing system 1 generates the result of the semantic analysis process for the selected sentence. For example, the information processing system 1 estimates information such as a domain goal of a selected sentence by a speech semantic analysis process, and generates information of a semantic frame including information such as a semantic analysis score indicating the certainty of the estimated information. ..

Then, the information processing system 1 performs analysis accuracy conversion based on the result of the utterance semantic analysis process and the score function (step S410). For example, the information processing system 1 performs analysis accuracy conversion using the equation (1). The information processing system 1 calculates the accuracy index value (semantic analysis accuracy) indicating the semantic analysis accuracy of the selected sentence by inputting the semantic analysis score and the information indicating the language of the selected sentence into the equation (1). Further, the information processing system 1 makes it possible to identify the sentence in the translated sentence list as a processed sentence by adding a flag to the sentence for which the calculation of the accuracy index value (semantic analysis accuracy) has been completed. The information processing system 1 may exclude sentences for which the accuracy index value (semantic analysis accuracy) has been calculated from the translated sentence list.

The information processing system 1 performs the processes of steps S409 to S410 by selecting each sentence in the translated sentence list by loop processing of the translated sentence list. As a result, the information processing system 1 calculates the accuracy index value (semantic analysis accuracy) of each sentence in the translated sentence list by the loop processing of the translated sentence list.

When there is an unprocessed sentence in the translated sentence list (step S411: No), the information processing system 1 selects one unprocessed sentence in the translated sentence list and performs the processes of steps S409 to S410.

On the other hand, the information processing system 1 ends the loop processing of the translated sentence list when there is no unprocessed sentence in the translated sentence list (step S411: Yes).

Then, the information processing system 1 selects a semantic frame (step S412). The information processing system 1 selects the sentence having the maximum accuracy index value (semantic analysis accuracy) from each sentence as the sentence to be used in the subsequent processing. The information processing system 1 selects the semantic analysis result of the sentence having the maximum accuracy index value (semantic analysis accuracy) among the sentences as the sentence to be used in the subsequent processing. The information processing system 1 selects, among the sentences, the semantic frame corresponding to the sentence having the maximum accuracy index value (semantic analysis accuracy) as the sentence to be used in the subsequent processing.

Then, the information processing system 1 performs slot inverse transformation (step S413). The information processing system 1 performs slot inverse conversion using a sentence meaning frame (semantic frame with maximum accuracy) having the maximum accuracy index value (semantic analysis accuracy). The information processing system 1 converts the slot value in the meaning frame having the maximum accuracy into the slot value of the input language (target language). The information processing system 1 converts the slot value of the specific language (translation destination language) into the slot value of the input language (target language). Note that the information processing system 1 does not have to perform step S413 when the inverse transformation is unnecessary, such as when the language of the sentence having the maximum accuracy index value (semantic analysis accuracy) is the target language.

Then, the information processing system 1 generates a response (step S414). The information processing system 1 generates information such as images, sounds, and texts according to the output mode. For example, the information processing system 1 generates a response using the result of the semantic analysis process of the sentence having the maximum accuracy index value (semantic analysis accuracy) of each sentence.

Further, when the information processing system 1 determines that the language is not the target language (step S403: No), the information processing system 1 rejects the process (step S415). For example, when the information processing system 1 determines that the language is not the target language, the information processing system 1 interrupts the process. Then, the information processing system 1 generates a response in step S415 using the reason for interruption. For example, the information processing system 1 uses the reason for interruption to generate a response such as "processing is interrupted because of a language other than the target".

Then, the information processing system 1 outputs the generated information (step S416). The information processing system 1 displays images and texts and outputs sounds.

[1-7. Conceptual diagram of processing by information processing system]
Here, with reference to FIG. 14, each function, hardware configuration, and processing in the information processing system 1 are conceptually shown. FIG. 14 is a conceptual diagram showing an example of processing by the information processing system. The system process PS1 shown in FIG. 14 shows an example of the process realized by the information processing system 1. For example, FIG. 14 shows a schematic configuration diagram of the information processing system 1. Speech expansion, accuracy conversion, score function, semantic frame language selection, language expansion translator, slot inverse conversion, and response generation in FIG. 14 are important points for improving the accuracy and support of semantic analysis, for example. , Speech expansion, precision conversion, score function, and semantic frame language selection are very important parts of the process.

The system process PS1 shown in FIG. 14 is a diagram conceptually showing each process from the input of the user's utterance to the output of the response, and the functions and hardware configurations for realizing each process. For example, each process shown in the system process PS1 is executed by the information processing apparatus 100. Further, for example, the language expansion translator in the system processing PS1 is realized by the function of the conversion unit 132 of the information processing device 100. For example, the utterance meaning analyzer in the system processing PS1 is realized by the function of the execution unit 133 of the information processing device 100. In the following, a case where the information processing system 1 performs processing will be described as an example, but the processing shown in FIG. 14 is performed by any of the information processing device 100 and the terminal device 10 included in the information processing system 1. May be good. The same points as those in FIGS. 11 to 13 will be omitted as appropriate.

System processing As shown in PS1, the information processing system 1 performs language identification processing for utterances by the user. Then, when the information processing system 1 can identify the language of the input utterance language (input language), the information processing system 1 develops the utterance sentence. For example, the information processing system 1 generates a utterance sentence list including character information corresponding to the user's utterance and a paraphrase of the character information by expanding the utterance sentence. For example, the information processing system 1 generates an utterance sentence list including the utterance development information A, the utterance development information B, the utterance development information C, and the utterance development information D in FIG. 6 by the utterance sentence expansion. In this case, the information processing system 1 includes the utterance development information A in which the user's utterance is converted into a character string, the utterance development information B which is a paraphrase of the utterance development information A, the utterance development information C, the utterance development information D, and the like. Generate an utterance list.

Then, the information processing system 1 generates a translation list by the language expansion translator. The language expansion translator may be a plurality of translators that translate (convert) an input language into each of a plurality of languages. In the example of FIG. 6, the language expansion translator is a first translator that translates Japanese into English, a second translator that translates Japanese into Spanish, a third translator that translates Japanese into English-French, and the like. It may be composed of a plurality of translators of. The information processing system 1 generates character information in which the utterance development sentence is converted into each language by inputting each sentence (utterance development sentence) in the utterance sentence list into the translator of each language.

For example, the information processing system 1 converts each sentence (speech expansion sentence) included in the utterance sentence list into a sentence of each translation destination language, thereby converting a sentence (translation sentence) of each translation destination language corresponding to each utterance development sentence. ) Is generated. The information processing system 1 generates a translation list including sentences (translated sentences) of each translation destination language corresponding to each utterance development sentence. For example, the information processing system 1 uses a translator of each language for each of the utterance development information A, the utterance development information B, the utterance development information C, and the utterance development information D in FIG. A translation of each language corresponding to each of the development information B, the utterance development information C, and the utterance development information D is generated. Specifically, the information processing system 1 lists each of the utterance development information A, the utterance development information B, the utterance development information C, and the utterance development information D in FIG. 6 in the translated text in FIG. Generate translations for each language. Taking English as an example, the information processing system 1 translates each of the utterance development information A, the utterance development information B, the utterance development information C, and the utterance development information D in FIG. 6 into English (first translator). ) Is used to generate four English sentences corresponding to each of the utterance development information A, the utterance development information B, the utterance development information C, and the utterance development information D. Further, when the language corresponding to the user's utterance is also a language (specific language) capable of semantic analysis processing, the information processing system 1 adds each sentence of the utterance sentence list to the translation list to generate a translation list. In the example of FIG. 6, the information processing system 1 generates a translation list including the utterance development information A, the utterance development information B, the utterance development information C, and the utterance development information D because Japanese is also a specific language in the example of FIG.

Then, the information processing system 1 estimates the quality of the translation result. For example, the information processing system 1 estimates the quality of the translated text in the translation list.

Further, when the information processing system 1 determines that at least one translated sentence has a high score, the information processing system 1 performs a semantic analysis process by the utterance semantic analyzer using the translated sentence having a high score. When the information processing system 1 determines that at least one of the translated sentences in the translation list is of high quality, the information processing system 1 performs a semantic analysis process by the utterance semantic analyzer using the translated sentence of high quality. In this case, if there is a language (specific language) in which all the translated sentences are determined to have a low score, the information processing system 1 generates a response indicating that the processing is interrupted for that language. That is, the information processing system 1 generates a response indicating that the processing is interrupted when it is determined that the quality of all the translated sentences translated into a specific language is low. In this case, the information processing system 1 generates a response such as "The translation accuracy from the XX language to the YY language is low, so the process is interrupted." For example, the information processing system 1 generates a response such as "The input language is interrupted because the translation accuracy from Korean to English for semantic analysis is low."

When the input language is a specific language, the information processing system 1 performs semantic analysis processing by the utterance meaning analyzer using the utterance development sentence of the input language and the translated sentence having a high score. Further, when the input language is a specific language, the information processing system 1 may perform semantic analysis processing by the utterance meaning analyzer using the utterance expansion sentence even when there is no translated sentence having a high score. ..

The information processing system 1 performs processing related to post-editing when it is determined that all languages (specific languages) have low scores. For example, when the information processing system 1 determines that the quality of all the translated sentences is low, the information processing system 1 performs a process related to post-editing. When the input language is a specific language, the information processing system 1 performs semantic analysis processing by the utterance semantic analyzer using the utterance expansion sentence even when it is determined that the quality of all the translated sentences is low. , It is not necessary to perform the processing related to post-editing.

For example, when the information processing system 1 determines that the input language is not a specific language and the translated sentences of all the languages have a low score, and the processing for the user's utterance does not require immediacy, the translation result is obtained. Executes the process of manually editing. The information processing system 1 causes a cloud worker to perform manual editing. For example, the information processing system 1 causes the cloud worker to manually edit the translation result by transmitting the utterance development sentence and the translated sentence to a device (terminal device 10 or the like) used by the cloud worker.

Further, the information processing system 1 rejects the processing when the post-editing processing time is long (long time). For example, the information processing system 1 rejects the process if the manual editing by a cloud worker or the like is not completed within a predetermined time. In this case, the information processing system 1 generates a response such as "The translation time from the XX language to the YY language is long, so the process is interrupted." For example, the information processing system 1 generates a response such as "The translation time from the input language Dutch to Japanese for semantic analysis is long, so the process is interrupted."

Further, when the post-editing processing time is short (short time), the information processing system 1 performs the semantic analysis processing by the utterance semantic analyzer using the translated sentence edited by hand.

As described above, the information processing system 1 outputs each sentence (semantic analysis target sentence) by performing semantic analysis processing on the utterance development sentence and the translated sentence having a high score in the translation list by the utterance semantic analyzer. To get. For example, the information processing system 1 obtains the result of the semantic analysis for each sentence to be analyzed. For example, the information processing system 1 generates information of a semantic frame corresponding to each semantic analysis target sentence by a semantic analysis process. For example, the information processing system 1 specifies the Domain-Goal (domain goal) corresponding to each semantic analysis target sentence by the semantic analysis process, and generates information of the semantic frame in which the slot value is set. Further, the information processing system 1 generates information of a semantic frame including a Domain-Goal (domain goal) corresponding to each semantic analysis target sentence and a score (semantic analysis score) indicating the certainty of the slot value.

Then, the information processing system 1 converts the semantic analysis score (analysis score) of each semantic analysis target sentence into the semantic analysis accuracy (%) which is an accuracy index value. The information processing system 1 calculates the semantic analysis accuracy (%) by using the semantic analysis score (analysis score) and the score function. The information processing system 1 calculates the semantic analysis accuracy by using the information indicating the language and the score function that outputs the semantic analysis accuracy by inputting the semantic analysis score (analysis score).

Then, the information processing system 1 selects the language of the semantic frame. For example, in the information processing system 1, when there are a plurality of semantic analysis target sentences for which utterance semantic analysis processing has been performed, the language of the semantic frame of the semantic analysis target sentence is selected by selecting the semantic analysis target sentence based on the semantic analysis accuracy. Is selected as the language to be used for the subsequent processing. The information processing system 1 selects the language of the semantic frame of the semantic analysis target sentence as the language to be used for the subsequent processing by selecting the semantic analysis target sentence having the maximum semantic analysis accuracy. When the information processing system 1 has one semantic analysis target sentence for which the utterance semantic analysis process has been performed, the information processing system 1 selects the language of the semantic frame of the semantic analysis target sentence as the language to be used for the subsequent processing.

Then, the information processing system 1 performs slot inverse transformation using the knowledge DB. The information processing system 1 performs reverse slot conversion using reverse translation or a multilingual dictionary. For example, the information processing system 1 performs slot inverse conversion using a knowledge database such as an external knowledge information providing server or a knowledge information storage unit 125. For example, the information processing system 1 reversely converts the slot value of the translation destination language in the semantic frame into the slot value of the input language (target language). The information processing system 1 does not have to perform the slot inverse transformation process when the inverse transformation is unnecessary, such as when the language of the semantic frame is the input language (target language).

Then, the information processing system 1 generates a response. The information processing system 1 generates information corresponding to the result of the semantic analysis process. The information processing system 1 generates response information based on the domain-goal (domain goal) of the semantic frame and the slot value information.

[1-7-1. Specific example of processing by information processing system]
A specific example of the process will be described with reference to FIG. FIG. 15 is a diagram showing an example of specific processing by the information processing system. For example, the processing example shown in FIG. 15 corresponds to the flow shown in FIG. The process shown in FIG. 15 may be performed by any of the information processing device 100 and the terminal device 10 included in the information processing system 1. The same points as those in FIGS. 11 to 14 will be omitted as appropriate.

In FIG. 15, the user speaks in Korean, saying "Play a song of XXX (Korean)". As a result, the information processing system 1 acquires the input IN1 of the user in Korean. It is assumed that "XXX" is a predetermined artist name (singer name).

Then, the information processing system 1 performs semantic analysis processing in Korean. As a result, the information processing system 1 acquires the result of the semantic analysis in Korean as shown in the result RS1. In the example of FIG. 15, the information processing system 1 acquires a result indicating that the domain is "cooking" and the score value is "low" in the semantic analysis in Korean. The information processing system 1 uses a Korean semantic analyzer to generate a semantic analysis result in which the domain is specified as "cooking" and the semantic analysis score is a low numerical value such as "0.2".

Since the score of the semantic analysis in Korean is low and the accuracy is low, the information processing system 1 performs the semantic analysis process using the English translation of the Korean word "Korean song (Korean)". conduct. The information processing system 1 performs a semantic analysis process using the English translation "Play XXX song", and as shown in the result RS2, the domain (domain goal) is "music playback" and its score. Get the result indicating that the value is "high". The information processing system 1 uses an English semantic analyzer to generate a semantic analysis result in which the domain is specified as "music reproduction" and the semantic analysis score is a high numerical value such as "0.9".

The information processing system 1 determines the response to the user using the English semantic analysis result. The information processing system 1 determines to play the song of the artist "XXX" using the English semantic analysis result. It is assumed that the song title is "AAA & BBB" artist "XXX". In FIG. 15, the information processing system 1 outputs a response in Korean saying "AAA & BBB is reproduced (Korean)". Then, the information processing system 1 executes a process of playing the song "AAA & BBB" of the artist "XXX".

The information processing system 1 may notify the user of the input language (Korean) and the analysis language (English). In addition, the information processing system 1 provides feedback to the user. As shown in the feedback FB1, the information processing system 1 notifies the user in Korean that "the accuracy of Korean is low, so the analysis was performed in English (Korean)". The notification mode may be various modes such as voice output and display on the screen. Further, the information processing system 1 may turn on a light of a color (for example, green or the like) indicating that the analysis is in English, or may output a sound effect (for example, pip or the like) meaning translation. ..

[1-8. Details of processing, etc.]
From here, the details of the various processes described above will be described.

[1-8-1. Language identification]
First, the above-mentioned language identification will be described. Language identification is a known technique that assumes a plurality of input languages and automatically identifies the language of the speaker, and is often used at the same time as voice recognition processing in recent years. In addition, language identification is roughly divided into a case of judging from an acoustic model of speech recognition and a case of judging from a language model. The former has the advantage that the time (delay) required for language identification can be reduced, and the latter enables more accurate language identification, which is a trade-off relationship between delay and accuracy. The information processing system 1 may perform language identification by either a process of determining from an acoustic model of voice recognition or a process of determining from a language model.

Also, if the input language is not supported by the translator or a specific language of the semantic analyzer, the processing is interrupted and the reason for the interruption is included in the response. For example, the information processing system 1 outputs such as "ZZ language does not support translation and semantic analysis".

[1-8-2. Improvement of low-precision semantic analyzer]
Next, the improvement of the low-precision semantic analyzer will be described. For example, the information processing system 1 processes "N × M" sentences obtained by multiplying the number of utterance developments "N" and the number of language developments "M".

Note that the information processing system 1 may first perform semantic analysis processing of only input utterance and input language when there is a viewpoint of time constraints, for example. In this case, the information processing system 1 makes it possible to execute either the utterance sentence expansion process, the language expansion process, or both when the result of the semantic analysis process of only the input utterance and the input language is low accuracy. For example, when the score value (semantic analysis score) or the semantic analysis accuracy of the semantic analysis processing of only the input utterance and the input language is less than a certain value (threshold), the information processing system 1 performs the processing of the utterance sentence expansion and the language expansion. Make either or both feasible. In this case, the lower the constant value (threshold value), the more inaccurate semantic analysis processing is allowed for the input content, but the value of this constant value (threshold value) can be set as appropriate. .. For example, the value of this constant value (threshold value) may be appropriately set according to the purpose of processing and the like. For example, the higher the accuracy required for processing, the larger the constant value (threshold value) may be set.

In the example of FIG. 14, the input language is Korean and the semantic analysis process becomes the cooking domain due to low accuracy, and then the utterance sentence converted by the English translation makes the semantic analysis process highly accurate again and is changed to the music domain. .. In this case, the information processing system 1 can include a response sentence such as "translated and analyzed in English because of low accuracy" in the response.

[1-8-3. Quality estimation]
Next, quality estimation will be described. Quality estimation is a process (module) for estimating the accuracy of the output of a translator. If the accuracy is low, it is difficult to obtain the desired result even with the semantic analysis process, and in order to avoid the disadvantage to the user due to the semantic analysis estimating another domain goal, at the stage of the translation result. This is to stop the processing. That is, the information processing system 1 performs a normal semantic analysis process when the score value (quality score) of the quality estimation is higher than a certain level, interrupts the process when the score value is lower than a certain level, and includes the reason for the interruption in the response. For example, when only the quality scores of English, French and German have a threshold value of "0.75", the information processing system 1 has three languages of English, French and German, which are languages having a threshold value of "0.75" or higher. Performs normal semantic analysis processing, other languages interrupt the processing and include the reason for the interruption in the response.

For example, in the case of text display or voice response, the information processing system 1 outputs such as "The processing is interrupted because the translation accuracy from Dutch to Japanese is low". This allows the user to know which language was unable to successfully translate. The information processing system 1 may output (notify) in various expression formats as long as it can be recognized by the user, such as displaying an error sound or an identifiable color on the device.

Further, the information processing system 1 calculates a quality estimation score value (quality score) for each sentence of all the translated sentence lists generated by the utterance sentence expansion and the language expansion, and any one of them is above a certain level. If it is a value, the semantic analysis process is executed for all specific languages above a certain level. On the other hand, when all the specific languages have values below a certain level, the information processing system 1 interrupts the processing and includes the reason for the interruption in the response. For example, the information processing system 1 outputs such as "Translation to a processable language has low accuracy, so processing is interrupted."

[1-8-4. Cloud worker]
Next, the cloud worker will be described. As an extension of the addition of quality estimation, the information processing system 1 causes a cloud worker (translation editor) to manually correct the translation result (post-editing) when the translation result of all specific languages is below a certain level. It is also possible. However, since it is manually edited and it takes a long time to process, this process is expected to be used for which immediacy is not required. Therefore, if the processing is not completed even after a certain period of time has passed, the information processing system 1 interrupts the processing and includes the reason for the interruption in the response. For example, the information processing system 1 outputs such as "It will be interrupted because the manual translation from Hindi to Japanese takes a long time." Further, the information processing system 1 may adjust a certain time or more according to the domain goal.

[1-8-5. Translation accuracy and time adjustment]
The information processing system 1 may change various values such as a threshold value. For example, in the information processing system 1, the user may freely set the translation accuracy and the waiting time above a certain level in addition to the default values of the system. Depending on the user, if you want to obtain only high quality products, set a high accuracy threshold, and if you want results even if it takes time, you can freely customize by setting a long waiting time. , It is possible to optimize for each user.

[2. Other embodiments]
The processing according to each of the above-described embodiments may be carried out in various different forms (modifications) in addition to the above-mentioned embodiments and modifications.

[2-1. Configuration example of performing semantic analysis processing on the client side]
In the embodiment, as an example of the system configuration, the case where the information processing apparatus 100 performs conversion processing, semantic analysis processing, inverse transformation theory, etc. is shown, but the terminal device 10 performs conversion processing, semantic analysis processing, inverse transformation theory, etc. May be done. That is, the terminal device 10 which is a device on the client side may be an information processing device that performs the above-mentioned conversion processing, semantic analysis processing, inverse conversion theory, and the like. As described above, the system configuration of the information processing system 1 is not limited to the configuration in which the information processing device 100, which is a device on the server side, performs conversion processing, semantic analysis processing, inverse transformation, and the like, and is a terminal device which is a device on the client side. 10 may be configured to perform the above-mentioned conversion processing, semantic analysis processing, inverse transformation theory, and the like.

When the terminal device 10 is an information processing device that performs the above-mentioned conversion processing, semantic analysis processing, inverse conversion theory, etc., the information processing system 1 performs translation, semantic analysis, and inverse conversion on the client side (terminal device 10). Then, the server side (information processing device 100) acquires the information of the semantic analysis result and the inverse conversion result from the terminal device 10 and performs various processes. In this case, the execution unit 152 of the terminal device 10 may have the same function as the execution unit 133 of the information processing device 100. Further, the terminal device 10 may have a conversion unit that realizes the same function as the conversion unit 132 described above, and an inverse conversion unit that realizes the same function as the inverse conversion unit 136. Further, in this case, the information processing apparatus 100 does not have to have the conversion unit 132 and the inverse conversion unit 136.

Further, the information processing system 1 may have a system configuration in which the client side (terminal device 10) analyzes the meaning of the utterance and the server side (information processing device 100) performs inverse conversion. In this case, the terminal device 10 which is a device on the client side is an information processing device which performs the above-mentioned conversion processing and the semantic analysis processing, and the information processing device 100 which is a device on the server side performs the above-mentioned inverse conversion processing. It may be. In this case, the conversion unit and the execution unit 152 of the terminal device 10 perform conversion processing and semantic analysis processing, and the inverse conversion unit 136 of the information processing device 100 performs reverse conversion processing.

Note that the above is an example, and in the information processing system 1, any device may perform each process. As described above, the information processing system 1 may have a system configuration in which either the client-side device (terminal device 10) or the server-side device (information processing device 100) performs each process.

[2-2. Other configuration examples]
In the above example, the case where the information processing device 100 and the terminal device 10 are separate bodies is shown, but these devices may be integrated.

[2-3. others]
Further, among the processes described in each of the above embodiments, all or a part of the processes described as being automatically performed can be manually performed, or the processes described as being manually performed. It is also possible to automatically perform all or part of the above by a known method. In addition, the processing procedure, specific name, and information including various data and parameters shown in the above document and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the illustrated information.

Further, each component of each device shown in the figure is a functional concept, and does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each device is not limited to the one shown in the figure, and all or part of the device is functionally or physically dispersed / physically distributed in any unit according to various loads and usage conditions. Can be integrated and configured.

Further, each of the above-described embodiments and modifications can be appropriately combined as long as the processing contents do not contradict each other.

Further, the effects described in the present specification are merely examples and are not limited, and other effects may be obtained.

[3. Effect of this disclosure]
As described above, the information processing device (information processing device 100 in the embodiment) according to the present disclosure includes an execution unit (execution unit 133 in the embodiment) and a calculation unit (calculation unit 134 in the embodiment). The execution unit executes a semantic analysis process on one or more character information corresponding to each of one or more languages including a target language which is a language corresponding to the user's utterance. The calculation unit is an accuracy index that makes it possible to compare the accuracy of the semantic analysis processing for each of the one or more character information among a plurality of languages based on the result of the semantic analysis processing corresponding to each of the one or more character information. Calculate the value.

As described above, the information processing apparatus according to the present disclosure determines the accuracy of the semantic analysis processing for each of the one or more character information between a plurality of languages based on the result of the semantic analysis processing of each of the one or more character information. By calculating the accuracy index value that can be compared with, it is possible to compare the semantic analysis accuracy between languages.

Further, the calculation unit calculates each accuracy index value of one or more character information by using a function that outputs the accuracy index value by inputting the information of the result of the semantic analysis process. As described above, the information processing apparatus can appropriately calculate the accuracy index value by using the function that outputs the accuracy index value by inputting the information of the result of the semantic analysis process. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

In addition, the calculation unit calculates the accuracy index value of each of one or more character information by using a function that outputs the accuracy index value by inputting the score included in the result of the semantic analysis process. As described above, the information processing apparatus can appropriately convert the score into the accuracy index value by using the function that outputs the accuracy index value by inputting the score included in the result of the semantic analysis process. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

In addition, the calculation unit uses a function that outputs the accuracy index value by inputting the score of the semantic analysis process for the character information of one language and the information indicating one language, and the accuracy index of the character information of one language. Calculate the value. As described above, the information processing apparatus can appropriately convert the score into the accuracy index value by using the function that outputs the accuracy index value by inputting the score and the language information. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

In addition, the execution unit executes a semantic analysis process on one or more character information including one character information corresponding to the user's utterance in the target language. In this way, the information processing device can compare the semantic analysis accuracy between languages including the target language by calculating the accuracy index value for each of one or more character information including the character information of the target language. Can be.

In addition, the execution unit executes a semantic analysis process on one or more character information including the translated character information in which one character information is converted into the translation destination language which is the translation destination of the target language. In this way, the information processing device calculates the accuracy index value for each of one or more character information including the character information converted into the translation destination language, thereby performing semantic analysis between the languages including the translation destination language. It is possible to compare the accuracy.

In addition, the execution unit executes a semantic analysis process on one or more character information including a paraphrase in which one character information of the target language is paraphrased into another expression of the target language. In this way, the information processing device calculates the accuracy index value for each of one or more character information including the paraphrase paraphrased into another expression of the character information of the target language, so that the user can make a plurality of utterances. After expanding to the expression of, it is possible to compare the accuracy of each of the expressions. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

In addition, the execution unit executes a semantic analysis process on one or more character information including the translation paraphrase in which the paraphrase of the target language is converted into the translation destination language to which the target language is translated. In this way, the information processing device sets the accuracy index value for each of one or more character information including the character information in which the paraphrase paraphrased into another expression of the character information of the target language is converted into the translation destination language. By calculating, it is possible to expand the user's utterance into a plurality of expressions, translate each expression, and then compare the accuracy of each expression. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

Further, the information processing device according to the present disclosure includes a conversion unit (conversion unit 132 in the embodiment). The conversion unit converts the character information corresponding to the target language into the character information corresponding to the translation destination language to be the translation destination of the target language. The execution unit executes a semantic analysis process on the character information converted by the conversion unit. In this way, the information processing device converts the character information corresponding to the target language into the character information corresponding to the translation destination language to be the translation destination of the target language, and performs semantic analysis processing to perform the semantic analysis of each translation destination language. The accuracy index value corresponding to can be calculated. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

In addition, the conversion unit converts the character information corresponding to the target language into the character information corresponding to the language whose meaning can be interpreted. In this way, the information processing device converts the character information corresponding to the target language into the character information corresponding to the language that can interpret the meaning (specific language), and performs the semantic analysis process to convert the character information into each translation destination language. The corresponding accuracy index value can be calculated. Therefore, the information processing device can make it possible to compare the accuracy of semantic analysis between languages.

Further, the information processing apparatus according to the present disclosure includes a selection unit (selection unit 135 in the embodiment). The selection unit selects the processing target character information, which is the character information used for processing, from the one or more character information based on each accuracy index value of the one or more character information calculated by the calculation unit. In this way, the information processing apparatus appropriately selects the character information used for processing by selecting the processing target character information which is the character information used for processing from among one or more character information based on the accuracy index value. can do. Therefore, the information processing apparatus can compare the accuracy of semantic analysis between languages and appropriately perform processing based on the comparison result.

In addition, the selection unit selects the character information having the maximum accuracy index value as the character information to be processed. As described above, the information processing apparatus can perform processing using the character information having the maximum accuracy by selecting the character information having the maximum accuracy index value as the character information to be processed. Therefore, the information processing apparatus can compare the accuracy of semantic analysis between languages and appropriately perform processing based on the comparison result.

In addition, the calculation unit calculates the accuracy index value corresponding to the character information of the target language. When the accuracy index value corresponding to the character information of the target language is equal to or higher than a predetermined value, the selection unit selects the character information of the target language as the processing target character information. In this way, when the accuracy index value corresponding to the character information of the target language is equal to or higher than a predetermined value, the information processing apparatus selects the character information of the target language as the processing target character information, so that the accuracy of the target language is high. In some cases, the language can be used for processing. Therefore, the information processing apparatus can appropriately perform processing using the target language.

Further, when the accuracy index value corresponding to the target language is equal to or higher than the predetermined value, the execution unit does not execute the semantic analysis process of a language other than the target language. As described above, when the accuracy index value corresponding to the target language is equal to or higher than the predetermined value, the information processing apparatus can suppress an increase in the processing load by not executing the semantic analysis process of a language other than the target language. .. Therefore, the information processing apparatus can appropriately perform processing while suppressing an increase in processing load.

Further, the information processing apparatus according to the present disclosure includes an inverse conversion unit (in the embodiment, an inverse conversion unit 136). The inverse transformation unit converts the result of the semantic analysis process corresponding to the language of the character information to be processed into the target language. In this way, the information processing apparatus can obtain the information of the semantic analysis corresponding to the target language by converting the result of the semantic analysis process corresponding to the language of the character information to be processed into the target language. Therefore, the information processing apparatus can appropriately perform processing using the target language.

In addition, the inverse conversion unit converts the result of the semantic analysis process into the target language when the language of the processing target character information is other than the target language. In this way, when the language of the character information to be processed is other than the target language, the information processing device converts the result of the semantic analysis process into the target language, and the result of the semantic analysis other than the target language is converted into the target language. You can get the information available in. Therefore, the information processing apparatus can appropriately perform processing using the target language.

In addition, the inverse transformation unit converts a part of the result of the semantic analysis process into the target language. In this way, the information processing device converts a part of the result of the semantic analysis process into the target language, and by converting only the necessary information into the target language, the result of the semantic analysis other than the target language is obtained. , You can get the information available in the target language. Therefore, the information processing apparatus can appropriately perform processing using the target language.

In addition, the inverse conversion unit converts the slot value of the result of the semantic analysis process into the target language. In this way, the information processing device converts the slot value of the result of the semantic analysis process into the target language, thereby converting only the information (slot value) necessary for executing the service, for example, into the target language. It is possible to obtain information that makes the results of semantic analysis in a language other than the target language available in the target language. Therefore, the information processing apparatus can appropriately perform processing using the target language.

Further, when the target language is language identifiable, the execution unit executes a semantic analysis process on one or more character information corresponding to each of the one or more languages including the target language. As described above, when the target language is language-identifiable, the information processing apparatus can appropriately perform processing on the language-identifiable language by executing the semantic analysis process. In addition, the information processing device can perform processing on a language that is not language-identifiable, and can suppress improper processing or output of inappropriate results.

[4. Hardware configuration]
Information devices such as the information processing device 100 and the terminal device 10 according to each of the above-described embodiments and modifications are realized by, for example, a computer 1000 having a configuration as shown in FIG. FIG. 20 is a hardware configuration diagram showing an example of a computer 1000 that realizes the functions of information processing devices such as the information processing device 100 and the terminal device 10. Hereinafter, the information processing apparatus 100 according to the embodiment will be described as an example. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM (Read Only Memory) 1300, an HDD (Hard Disk Drive) 1400, a communication interface 1500, and an input / output interface 1600. Each part of the computer 1000 is connected by a bus 1050.

The CPU 1100 operates based on the program stored in the ROM 1300 or the HDD 1400, and controls each part. For example, the CPU 1100 expands the program stored in the ROM 1300 or the HDD 1400 into the RAM 1200 and executes processing corresponding to various programs.

The ROM 1300 stores a boot program such as a BIOS (Basic Input Output System) executed by the CPU 1100 when the computer 1000 is started, a program that depends on the hardware of the computer 1000, and the like.

The HDD 1400 is a computer-readable recording medium that non-temporarily records a program executed by the CPU 1100 and data used by the program. Specifically, the HDD 1400 is a recording medium for recording an information processing program according to the present disclosure, which is an example of program data 1450.

The communication interface 1500 is an interface for the computer 1000 to connect to an external network 1550 (for example, the Internet). For example, the CPU 1100 receives data from another device or transmits data generated by the CPU 1100 to another device via the communication interface 1500.

The input / output interface 1600 is an interface for connecting the input / output device 1650 and the computer 1000. For example, the CPU 1100 receives data from an input device such as a keyboard or mouse via the input / output interface 1600. Further, the CPU 1100 transmits data to an output device such as a display, a speaker, or a printer via the input / output interface 1600. Further, the input / output interface 1600 may function as a media interface for reading a program or the like recorded on a predetermined recording medium (media). The media is, for example, an optical recording medium such as DVD (Digital Versatile Disc) or PD (Phase change rewritable Disk), a magneto-optical recording medium such as MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. Is. For example, when the computer 1000 functions as the information processing device 100 according to the embodiment, the CPU 1100 of the computer 1000 realizes the functions of the control unit 130 and the like by executing the information processing program loaded on the RAM 1200. Further, the information processing program according to the present disclosure and the data in the storage unit 120 are stored in the HDD 1400. The CPU 1100 reads the program data 1450 from the HDD 1400 and executes the program, but as another example, these programs may be acquired from another device via the external network 1550.

The present technology can also have the following configurations.
(1)
An execution unit that executes semantic analysis processing for one or more character information corresponding to each of one or more languages including a target language that is a language corresponding to a user's utterance.
An accuracy index that makes it possible to compare the accuracy of the semantic analysis process between a plurality of languages for each of the one or more character information based on the result of the semantic analysis process corresponding to each of the one or more character information. The calculation unit that calculates the value and
Information processing device equipped with.
(2)
The calculation unit
Using a function that outputs the accuracy index value by inputting the information of the result of the semantic analysis process, the accuracy index value of each of the one or more character information is calculated.
The information processing device according to (1).
(3)
The calculation unit
Using the function that outputs the accuracy index value by inputting the score included in the result of the semantic analysis process, the accuracy index value of each of the one or more character information is calculated.
The information processing device according to (2).
(4)
The calculation unit
Using the function that outputs the accuracy index value by inputting the score of the semantic analysis process for the character information of one language and the information indicating the one language, the accuracy of the character information of the one language. Calculate the index value,
The information processing device according to (3).
(5)
The execution unit
The semantic analysis process is executed on the one or more character information including one character information corresponding to the utterance of the user in the target language.
The information processing device according to any one of (1) to (4).
(6)
The execution unit
The semantic analysis process is executed on the one or more character information including the translated character information in which the one character information is converted into the translation destination language to be the translation destination of the target language.
The information processing device according to (5).
(7)
The execution unit
The semantic analysis process is executed on the one or more character information including the paraphrase in which the one character information of the target language is paraphrased into another expression of the target language.
The information processing device according to (5) or (6).
(8)
The execution unit
The semantic analysis process is executed on the one or more character information including the translation paraphrase in which the paraphrase of the target language is converted into the translation destination language to be the translation destination of the target language.
The information processing device according to (7).
(9)
A conversion unit that converts the character information corresponding to the target language into the character information corresponding to the translation destination language to be the translation destination of the target language.
With more
The execution unit
The semantic analysis process is executed on the character information converted by the conversion unit.
The information processing device according to any one of (6) to (8).
(10)
The conversion unit
The character information corresponding to the target language is converted into the character information corresponding to the language that can interpret the meaning.
The information processing device according to (9).
(11)
A selection unit that selects processing target character information, which is character information used for processing, from the one or more character information based on the accuracy index value of each of the one or more character information calculated by the calculation unit.
The information processing apparatus according to any one of (1) to (10).
(12)
The selection unit
The character information having the maximum accuracy index value is selected as the processing target character information.
The information processing device according to (11).
(13)
The calculation unit
The accuracy index value corresponding to the character information of the target language is calculated, and the accuracy index value is calculated.
The selection unit
When the accuracy index value corresponding to the character information of the target language is equal to or higher than a predetermined value, the character information of the target language is selected as the processing target character information.
The information processing device according to (11) or (12).
(14)
The execution unit
When the accuracy index value corresponding to the target language is equal to or higher than a predetermined value, the semantic analysis process of a language other than the target language is not executed.
The information processing device according to (13).
(15)
An inverse conversion unit that converts the result of the semantic analysis process corresponding to the language of the character information to be processed into the target language.
The information processing apparatus according to any one of (11) to (14).
(16)
The inverse conversion unit
When the language of the processing target character information is other than the target language, the result of the semantic analysis processing is converted into the target language.
The information processing device according to (15).
(17)
The inverse conversion unit
A part of the result of the semantic analysis process is converted into the target language.
The information processing apparatus according to (15) or (16).
(18)
The inverse conversion unit
Of the results of the semantic analysis process, the slot value is converted into the target language.
The information processing device according to any one of (15) to (17).
(19)
The execution unit
When the target language is language identifiable, the semantic analysis process is executed for the one or more character information corresponding to each of the one or more languages including the target language.
The information processing device according to any one of (1) to (18).
(20)
Semantic analysis processing is executed for one or more character information corresponding to each of one or more languages including the target language which is the language corresponding to the user's utterance.
An accuracy index that makes it possible to compare the accuracy of the semantic analysis process between a plurality of languages for each of the one or more character information based on the result of the semantic analysis process corresponding to each of the one or more character information. Calculate the value,
An information processing method that executes processing.

1 Information processing system 100 Information processing device 110 Communication unit 120 Storage unit 121 Language information storage unit 122 Semantic frame information storage unit 123 Analysis accuracy information storage unit 124 Threshold information storage unit 125 Knowledge information storage unit 130 Control unit 131 Acquisition unit 132 Conversion unit 133 Execution unit 134 Calculation unit 135 Selection unit 136 Inverse conversion unit 137 Generation unit 138 Transmission unit 10 Terminal device 11 Communication unit 12 Input unit 13 Output unit 14 Storage unit 15 Control unit 151 Reception unit 152 Execution unit 153 Reception unit 154 Transmission unit 16 Sensor unit 17 Display unit

Claims

An execution unit that executes semantic analysis processing for one or more character information corresponding to each of one or more languages including a target language that is a language corresponding to a user's utterance.
An accuracy index that makes it possible to compare the accuracy of the semantic analysis process between a plurality of languages for each of the one or more character information based on the result of the semantic analysis process corresponding to each of the one or more character information. The calculation unit that calculates the value and
Information processing device equipped with.
The calculation unit
Using a function that outputs the accuracy index value by inputting the information of the result of the semantic analysis process, the accuracy index value of each of the one or more character information is calculated.
The information processing device according to claim 1.
The calculation unit
Using the function that outputs the accuracy index value by inputting the score included in the result of the semantic analysis process, the accuracy index value of each of the one or more character information is calculated.
The information processing device according to claim 2.
The calculation unit
Using the function that outputs the accuracy index value by inputting the score of the semantic analysis process for the character information of one language and the information indicating the one language, the accuracy of the character information of the one language. Calculate the index value,
The information processing device according to claim 3.
The execution unit
The semantic analysis process is executed on the one or more character information including one character information corresponding to the utterance of the user in the target language.
The information processing device according to claim 1.
The execution unit
The semantic analysis process is executed on the one or more character information including the translated character information in which the one character information is converted into the translation destination language to be the translation destination of the target language.
The information processing device according to claim 5.
The execution unit
The semantic analysis process is executed on the one or more character information including the paraphrase in which the one character information of the target language is paraphrased into another expression of the target language.
The information processing device according to claim 5.
The execution unit
The semantic analysis process is executed on the one or more character information including the translation paraphrase in which the paraphrase of the target language is converted into the translation destination language to be the translation destination of the target language.
The information processing device according to claim 7.
A conversion unit that converts the character information corresponding to the target language into the character information corresponding to the translation destination language to be the translation destination of the target language.
With more
The execution unit
The semantic analysis process is executed on the character information converted by the conversion unit.
The information processing device according to claim 6.
The conversion unit
The character information corresponding to the target language is converted into the character information corresponding to the language that can interpret the meaning.
The information processing device according to claim 9.
A selection unit that selects processing target character information, which is character information used for processing, from the one or more character information based on the accuracy index value of each of the one or more character information calculated by the calculation unit.
The information processing apparatus according to claim 1.
The selection unit
The character information having the maximum accuracy index value is selected as the processing target character information.
The information processing device according to claim 11.
The calculation unit
The accuracy index value corresponding to the character information of the target language is calculated, and the accuracy index value is calculated.
The selection unit
When the accuracy index value corresponding to the character information of the target language is equal to or higher than a predetermined value, the character information of the target language is selected as the processing target character information.
The information processing device according to claim 11.
The execution unit
When the accuracy index value corresponding to the target language is equal to or higher than a predetermined value, the semantic analysis process of a language other than the target language is not executed.
The information processing device according to claim 13.
An inverse conversion unit that converts the result of the semantic analysis process corresponding to the language of the character information to be processed into the target language.
11. The information processing apparatus according to claim 11.
The inverse conversion unit
When the language of the processing target character information is other than the target language, the result of the semantic analysis processing is converted into the target language.
The information processing device according to claim 15.
The inverse conversion unit
A part of the result of the semantic analysis process is converted into the target language.
The information processing apparatus according to claim 15 or 16.
The inverse conversion unit
Of the results of the semantic analysis process, the slot value is converted into the target language.
The information processing device according to claim 15.
The execution unit
When the target language is language identifiable, the semantic analysis process is executed for the one or more character information corresponding to each of the one or more languages including the target language.
The information processing device according to claim 1.
Semantic analysis processing is executed for one or more character information corresponding to each of one or more languages including the target language which is the language corresponding to the user's utterance.
An accuracy index that makes it possible to compare the accuracy of the semantic analysis process between a plurality of languages for each of the one or more character information based on the result of the semantic analysis process corresponding to each of the one or more character information. Calculate the value,
An information processing method that executes processing.