JP7058574B2 - Information processing equipment, information processing methods, and programs - Google Patents

Description

The present invention relates to an information processing apparatus, an information processing method, and a program.

A technique for performing speech recognition using a latent language model is known (see Patent Document 1). The latent language model is a model that considers the latent words for each word in the learning text (corpus).

Japanese Patent No. 5975938

However, in the conventional technique, it is necessary to obtain the probabilities of tens of thousands of latent word candidates for each word in the corpus, and when the vocabulary is large, the processing load becomes high and it takes time to output the speech recognition result. May be required.

The present invention has been made in consideration of such circumstances, and one of the objects of the present invention is to provide an information processing device, an information processing method, and a program capable of performing voice recognition processing more efficiently. ..

One aspect of the present invention includes an acquisition unit that acquires voice data, an analysis unit that analyzes the voice data and converts it into text, and outputs one or more analysis results, and the input text related to the analysis result. A vector conversion unit that converts each of a plurality of words into a vector value by a distributed expression, a vector value converted by the vector conversion unit, and an input text whose input intention is known are obtained in advance. Based on the vector value of the known input text, the analysis result having a high possibility of reflecting the input intention of the input text of the user who emitted the voice related to the voice data is selected from the analysis results of 1 or more. It is an information processing apparatus including a selection unit.

According to one aspect of the present invention, the voice recognition process can be performed more efficiently.

It is a figure which shows an example of the use environment of the information processing apparatus 100 which concerns on embodiment. It is a figure which shows typically the process of the information processing apparatus 100. It is a block diagram of the information processing apparatus 100 which concerns on embodiment. It is a figure for demonstrating the vector conversion process by the W2V execution unit 106. It is a figure for demonstrating a sentence vector. It is a figure which shows typically the suitable candidate selection by a sorting unit 110. It is a figure for explaining a task text. It is a figure for demonstrating the reliability derivation process by a reliability derivation unit 110a. It is a figure which shows typically the task text vector list 120g. It is a figure for demonstrating a representative vector. It is a figure for demonstrating the similarity evaluation method. It is a figure which shows schematically the cluster selection by a language model arithmetic unit 112. It is a flowchart which shows an example of the flow of the language model generation processing by an information processing apparatus 100. It is a flowchart which shows an example of the flow of the voice recognition processing by an information processing apparatus 100.

Hereinafter, embodiments of the information processing apparatus, information processing method, and program of the present invention will be described with reference to the drawings.

[overview]
The information processing device is realized by one or more processors. The information processing device is a device that receives voice data in which voices emitted by a user are recorded, performs voice recognition processing of the received input data, and performs various processing based on the recognition result. Various processes include controlling the IoT (Internet of Things) device according to the intention of the user who emitted the voice, and responding to the user's question. Hereinafter, the operation of the information processing device intended by the user may be referred to as a task. The voice data may be compressed or encrypted.

FIG. 1 is a diagram showing an example of a usage environment of the information processing apparatus 100 according to the embodiment. In the illustrated environment, the terminal device 20, the controlled device 30, and the service server 40 communicate with each other via the network NW. The network NW includes, for example, a part or all of a WAN (Wide Area Network), a LAN (Local Area Network), the Internet, a provider device, a wireless base station, a dedicated line, and the like. In the example shown in FIG. 1, the number of controlled devices 30 is N (N is an integer of 1 or more). In this specification, when the individual control target devices 30-1 to 30-N are not distinguished, for example, when common matters are explained in the control target devices 30-1 to 30-N, the control target is simply controlled. Called device 30.

The terminal device 20 is a device that accepts a user's voice input. The terminal device 20 is a mobile phone such as a smartphone, a tablet terminal, a personal computer, a smart speaker (AI speaker), or the like.

The control target device 30 is an IoT device having a communication function and an interface for receiving control from the outside, and can be controlled in response to a request from the terminal device 20 operated by the user. The controlled device 30 is, for example, a television, a radio, a lighting fixture, a refrigerator, a microwave oven, a washing machine, a rice cooker, a self-propelled vacuum cleaner, an air conditioner, or the like.

The service server 40 communicates with a web server device that provides a web page corresponding to a request from a terminal device 20 operated by a user, and a terminal device 20 in which an application is started, and exchanges various information to provide contents. An application server device or the like that provides.

FIG. 2 is a diagram schematically showing the processing of the information processing apparatus 100. The information processing device 100 converts voice data input by the user via the terminal device 20 into a phoneme by applying it to an acoustic model, and uses one or more candidate texts (sounds included in the voice data) based on the phonemes. A suitable candidate is selected by generating a text version) and applying the candidate text selected based on the comparison with the known task features among the generated candidate texts to the language model. A suitable candidate is a candidate that is determined to be a suitable candidate that is likely to reflect the user's intention in the candidate text.

The acoustic model is a model for statistically analyzing frequency components and time changes to determine what kind of phonemes the input voice data is composed of (what is said). A phoneme is a label for specifying the smallest unit of a language such as an alphabet or a kana, and includes, for example, a vowel or a consonant. The information processing apparatus 100 obtains candidate texts by appropriately combining phonemes according to language rules.

As shown in FIG. 2, when the candidate text generated as a result of phoneme conversion is “kyonotenki”, for example, “k” and “t” indicate phonemes included in the generated candidate text. When the voice recognition process is performed on the premise of Japanese, the candidate text may be expressed in alphabetical notation, hiragana notation, or katakana notation. In the example shown in FIG. 2, the information processing apparatus 100 generates a candidate text including "kyonotenki", "kyonotenkii", and "kyonodenki" based on the received voice data.

In the example shown in FIG. 2, the information processing apparatus 100 performs morphological analysis on each of the conversion candidates including "kyonotenki", "kyonotenkii", and "kyonodenki". Morphological analysis is a process of determining a delimiter of words constituting a candidate text and deriving, for example, a part of speech of each delimited word. The morphological analysis is performed using, for example, a morphological analysis engine such as MeCAB. As a result of analyzing the candidate text "kyonotenki", for example, the information processing apparatus 100 derives three words "today (kyo)", "no (no)", and "weather (tenki)". Similarly, as a result of analyzing the candidate text "kyonotenkii", "today (kyo)", "(no)", "tenkii", and as a result of analyzing the candidate text "kyonodenki", "kyo (kyo)" , "(No)", "Electricity (denki)" are generated.

The information processing apparatus 100 evaluates the analysis result generated from each of the one or more candidate texts. Then, one candidate text is selected based on the evaluation value, and more specifically, the information processing apparatus 100 evaluates the matching rate of the analysis result of the candidate text with the feature amount obtained from the known task voice. , Select suitable candidates that are presumed to be in line with the user's intention. Then, the information processing apparatus 100 outputs an instruction regarding a task for generating output information corresponding to the intention. The precision rate will be described later.

[Constitution]
FIG. 3 is a configuration diagram of the information processing apparatus 100. The information processing apparatus 100 includes, for example, an acquisition unit 102, an analysis unit 104, a W2V (Word2Vec) execution unit 106, a text vector generation unit 108, a selection unit 110, a language model calculation unit 112, and a selection unit 114. , An output information generation unit 116, an output unit 118, and a storage unit 120. These components are realized by, for example, a hardware processor such as a CPU (Central Processing Unit) executing a program (software). In addition, some or all of these components are hardware (circuits) such as LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), and GPU (Graphics Processing Unit). It may be realized by the part; including circuitry), or it may be realized by the cooperation of software and hardware. The program may be stored in advance in a storage device (a storage device including a non-transient storage medium) such as an HDD or a flash memory of the information processing device 100, or a detachable storage such as a DVD or a CD-ROM. It is stored in a medium (non-transient storage medium), and may be installed in the HDD or flash memory of the information processing device 100 by mounting the storage medium in the drive device.

The storage unit 120 is realized by, for example, a RAM (Random Access Memory), a register, a flash memory, an EEPROM (Electrically Erasable Programmable Read Only Memory), or the like. The storage unit 120 is, for example, an acoustic model 120a, a language model 120b, a corpus analysis result 120c, a task text analysis result 120d, a candidate text analysis result 120e, a word vector list 120f, a task text vector list 120g, and a language model calculation. Information such as text 120h is stored.

The acquisition unit 102 acquires character information (for example, article data such as news and posted data of SNS (Social Networking Service)) used as a corpus in the information processing device 100 to perform voice recognition processing, and the analysis unit 104. Output to. The character information used as a corpus is colloquial text (for example, posting history in SNS, conversation history between the user and the device in the automatic response device, actual conversation converted into text, voice processing of the own device). History etc.) should be included. Further, the acquisition unit 102 acquires a data set of character information (hereinafter, task text) indicating a routine task set by the administrator of the information processing apparatus 100, and outputs the data set to the analysis unit 104.

Further, the acquisition unit 102 acquires the voice data input by the user of the terminal device 20 and outputs it to the analysis unit 104. When the user's position information is added to the voice data acquired by the acquisition unit 102, the position information shall be managed together with the candidate text.

The analysis unit 104 analyzes the character information used as the corpus output by the acquisition unit 102 by a predetermined analysis method. The predetermined analysis method is, for example, morphological analysis. Morphological analysis is the decomposition of character information into parts of speech such as nouns, verbs, and particles. The analysis unit 104 stores the analysis result in the storage unit 120 as the analysis result 120c of the corpus. Further, the analysis unit 104 analyzes the task text output by the acquisition unit 102, and stores the analysis result in the storage unit 120 as the analysis result 120d of the task text.

Further, the analysis unit 104 applies the voice data output by the acquisition unit 102 to the acoustic model 120a to generate one or more candidate texts, and then performs analysis processing such as morphological analysis on each candidate text. Further, the analysis unit 104 stores the analysis result in the storage unit 120 as the analysis result 120e of the candidate text.

FIG. 4 is a diagram for explaining the vector conversion process by the W2V execution unit 106. The W2V execution unit 106 generates a word vector by vector-expressing (distributed representation) the meaning of each word included in the analysis result 120c of the corpus, for example. In the example of FIG. 4, the W2V execution unit 106 generates the word vector of “volume”. The W2V execution unit 106 generates a word vector so that the distance (cosine similarity) between word vectors is close between words having similar meanings such as "sound" and "volume", and "music" and "music". do. The W2V execution unit 106 stores the generated vector value in the storage unit 120 as a word vector list 120f. The W2V execution unit 106 is an example of a “vector conversion unit”.

Further, when the task text or the candidate text contains a word that is not stored in the word vector list 120f, the W2V execution unit 106 adds the analysis result 120d of the task text or the analysis result 120e of the candidate text to, for example, the corpus. It may be analyzed in the same manner to generate those vector values. This vector value may or may not be stored in the word vector list 120f.

Returning to FIG. 3, the text vector generation unit 108 uses the vector values of the task text analysis result 120d, the candidate text analysis result 120e, and the word vector list 120f, and the vector value of the candidate text in sentence units (hereinafter, sentence). Vector) is generated. The text vector generation unit 108 outputs the generated sentence vector to the selection unit 110.

FIG. 5 is a diagram for explaining a sentence vector. For example, when the text vector generation unit 108 generates a sentence vector of "lower volume", the text vector generation unit 108 performs a predetermined operation on the word vectors of "volume", "o", and "lower" (for example,). (By adding each word vector), a sentence vector is generated. As a result, for the sentence vector that is the sum of the word vectors of the words that make up the sentence, the distance between the sentence vectors that have similar meanings such as "make the music sound quieter" and "turn down the volume". Will be closer.

Further, the text vector generation unit 108 generates a sentence vector of the task text using the analysis result 120d of the task text and the word vector output by the W2V execution unit 106, and stores it in the storage unit 120 as the task text vector list 120g. Let me. The task text is a text known to include the intention of the user, and is set in advance by, for example, the administrator of the information processing apparatus 100.

Returning to FIG. 3, the selection unit 110 selects the sentence vector that is the source of the language model 120b based on the sentence vector of the candidate text, the sentence vector of the task text, and the sentence vector of the language model calculation text 120h. The sorting unit 110 outputs the sorting result to the language model calculation unit 112. The language model calculation text 120h is, for example, a sentence vector of a task text assumed by the administrator of the information processing device 100 or a sentence vector held as a voice recognition processing history of the past information processing device 100.

The sorting unit 110 includes, for example, a reliability deriving unit 110a. The priority derivation process by the reliability derivation unit 110a will be described later.

The language model calculation unit 112 includes, for example, a language model generation unit 112a. The language model generation unit 112a generates a language model to which the selection result output by the selection unit 110 is applied, and stores it in the storage unit 120 as a language model 120b for each corpus. The language model generation unit 112a generates, for example, a language model 120h based on a language model calculation text 120h preset by the administrator of the information processing apparatus 100 and conversion candidates selected by the selection unit 110.

Further, the language model calculation unit 112 applies the candidate text output by the selection unit 110 to the language model 120b, and outputs the application result to the selection unit 114.

The selection unit 114 evaluates the candidate text output by the language model calculation unit 112 based on the evaluation value, and selects a suitable candidate having a high possibility of reflecting the input intention of the user. The selection unit 114 outputs a suitable candidate as a selection result to the output information generation unit 116.

When the position information is given to the candidate text, the selection unit 114 estimates the input environment of the user from the position information, evaluates whether or not the candidate text user includes the task execution intention, and evaluates. Candidate text may be selected based on the evaluation result. For example, when the user is estimated to be at home from the position information of the candidate text, the selection unit 114 sets a high accuracy rate of the task related to the controlled device 30 to be used at home, and at the same time, uses it at work. The accuracy with which the corresponding task is selected may be changed by setting the matching rate of the task with respect to the controlled device 30 to be low.

FIG. 6 is a diagram schematically showing suitable candidate selection by the selection unit 114. The language model is a model for generating suitable candidates from candidate texts. For example, from the similarity between the sentence vector of the candidate vector and the sentence vector of the task text, the selection unit 110 gives a higher evaluation value as the one is closer to the task text, and further, the score regarding the sequence of words is higher by using the language model. Suitable candidates are selected by comprehensively evaluating these evaluation values, which give higher evaluation values. The language model may be evaluated in consideration of the surrounding environment of the user.

Returning to FIG. 3, the output information generation unit 116 generates output information intended by the user based on the suitable candidates output by the selection unit 114, and outputs the output information to the output unit 118. The output information includes information for specifying the output destination device, a processing request for the output destination device, and the like.

For example, when the suitable candidate is "tell me the weather today", the output information generation unit 116 transmits a request to the weather forecast website provided by the service server 40, and transmits the request to the terminal device 20. The output information is the information including a part or all of the response of the request of. Further, for example, when the suitable candidate is "lower the volume of music", the output information generation unit 116 identifies the controlled target device 30 during music reproduction and outputs an instruction to lower the volume. When the output destination generates the output information of the control target device 30, the output information generation unit 116 also generates the output information notifying the terminal device 20 that the output information has been output to the control target device 30. You may.

The output unit 118 outputs the output information output by the output information generation unit 116 to the terminal device 20 or the controlled device 30.

[Task text]
The task text will be described below. The administrator of the information processing apparatus 100 extracts, for example, a task text as an evaluation standard by the selection unit 114 based on the past voice input history of the terminal apparatus 20 and the processing history of the information processing apparatus 100.

FIG. 7 is a diagram for explaining the task text. The left figure of FIG. 7 shows the voice recognition results R1 to R7 of the past voice input history of the terminal device 20. The voice recognition result includes a result reflecting the input intention of the user of the terminal device 20 and a voice recognition result having no input intention by the user. For example, when the administrator of the information processing apparatus 100 determines that the voice recognition result R4 is a text close to the task, the administrator sets a high priority as shown in the upper right figure of FIG. Close to a task means that it contains text that is likely to reflect the user's input intent. Further, when the administrator of the information processing apparatus 100 determines that the voice recognition result R6 is a text far from the task, the administrator sets the priority low as shown in the lower right figure of FIG.

Further, the administrator of the information processing apparatus 100 determines that the texts R1, R2, R3, R5, and R7 are far from the task, and sets the priority low. The priority of the task text is registered in the task text vector list 120g together with the sentence vector value of the task text, for example.

[Language model generation process flow]
Hereinafter, the process of generating the language model 120b by the information processing apparatus 100 will be described. The information processing apparatus 100 generates, for example, a language model 120b for each type of corpus. Further, the administrator of the information processing apparatus 100 may periodically change / update the language model calculation text 120h, and for example, the language model is regenerated at that timing.

FIG. 8 is a flowchart showing an example of the flow of the generation processing of the language model 120b by the information processing apparatus 100.

First, the acquisition unit 102 acquires character information to be used as a corpus (S100). Next, the analysis unit 104 analyzes the character information used as the corpus, and stores the analysis result in the storage unit 120 as the analysis result 120c of the corpus (S102). Next, the W2V execution unit 106 generates a vector value of the word included in the analysis result 120c of the corpus and stores it in the storage unit 120 as the word vector list 120f (S104).

Next, the acquisition unit 102 acquires the task text (S106). Next, the analysis unit 104 analyzes the task text and stores the analysis result in the storage unit 120 as the analysis result 120d of the task text (S108).

Next, the acquisition unit 102 acquires the candidate text (S110). Next, the analysis unit 104 analyzes the candidate text and stores the analysis result in the storage unit 120 as the analysis result 120e of the candidate text (S112).

Next, the text vector generation unit 108 generates a sentence vector of the task text by referring to the analysis result 120d of the task text and the word vector list 120f, and stores it in the storage unit 120 as the task text vector list 120g (S114). ..

Next, the selection unit 110 selects the candidate text and outputs it to the language model generation unit 112a (S116).

Next, the language model generation unit 112a generates a language model 120b based on the candidate text output by the selection unit 110 and the language model calculation text 120h (S120). This is the end of the description of the processing of this flowchart.

[Degree of reliability]
Hereinafter, the reliability derivation process of the reliability derivation unit 110a will be described more specifically. The reliability is a numerical value between 0 and 1.0 indicating the degree of evaluation of the reliability of the speech recognition result. The reliability derivation unit 110a sets the reliability to 1.0, for example, when the reliability of the text is high, that is, when there is no other text that can be a competitor. The reliability is derived, for example, using the posterior probabilities of words obtained as search results of a large vocabulary continuous speech recognition engine.

FIG. 9 is a diagram for explaining the reliability derivation process by the reliability derivation unit 110a. The reliability derivation unit 110a derives, for example, the reliability of each of the candidate texts E1 to E4. The sorting unit 110 selects, for example, the candidate texts E1 and E4 whose reliability derived by the reliability deriving unit 110a is equal to or higher than a threshold value (for example, about 0.8) as task texts. When a plurality of task texts can be selected, the sorting unit 110 may preferentially select a highly reliable task text.

[Vector list clustering]
FIG. 10 is a diagram schematically showing a task text vector list 120 g. The task text vector list 120g has, for example, a cluster structure of about 10. Organize similar task texts into a cluster. The cluster is configured by, for example, the k-means clustering method or the like.

Further, in the task text vector list 120g, the search efficiency can be improved by deriving the representative vector for each cluster. The representative vector may be, for example, the average of the sentence vectors of the task texts constituting the cluster, or may be the weighted average of the priority of the task text and the sentence vector.

FIG. 11 is a diagram for explaining a representative vector. When selecting a task text, the sorting unit 110 first compares the representative vector with the sentence vector of the candidate text to select a cluster, and then selects a suitable task text from the selected clusters.

[Similarity evaluation of text]
Hereinafter, a method for evaluating the similarity of texts will be described. FIG. 12 is a diagram for explaining a similarity evaluation method.

The language model calculation unit 112 applies, for example, the sentence vector v1 of "lowering the volume" and the sentence vector v2 of "lowering the sound of music" to the mathematical formula for obtaining the cosine similarity shown in the equation (1). By doing so, the similarity of the text is evaluated.

Equation (1) is an equation representing dividing the product of the sentence vector v1 and the sentence vector v2 by the product of the absolute value of the sentence vector v1 and the absolute value of the sentence vector v2. It shows that the vector v1 and the sentence vector v2 are similar.

If the cosine similarity is equal to or higher than the threshold value, the language model calculation unit 112 determines that the sentence vector v1 and the sentence vector v2 are similar, that is, the original texts indicate the same or similar input intentions.

The language model calculation unit 112, for example, evaluates the similarity between the representative vector of the cluster and the sentence vector of the candidate text. FIG. 13 is a diagram schematically showing cluster selection by the language model calculation unit 112.

As shown in FIG. 13, the language model calculation unit 112 derives, for example, the similarity between the sentence vector of the candidate text “I want you to lower the volume” and the representative vectors of the clusters C1 and C2, respectively, and the similarity is determined. The high cluster C2 is selected as the selection target for the first stage. Further, the language model calculation unit 112 selects one or more suitable task texts as the selection target of the second stage from the selected cluster C2.

[Voice recognition processing]
FIG. 14 is a flowchart showing an example of the flow of voice recognition processing by the information processing apparatus 100.

First, the acquisition unit 102 acquires voice data (S200). Next, the analysis unit 104 applies the voice data output by the acquisition unit 102 to the acoustic model 120a to generate a candidate text (S202). Next, the language model calculation unit 112 applies the candidate text output by the analysis unit 104 to the language model 120b (S204). Next, the selection unit 114 selects a suitable candidate from the application result output by the language model calculation unit 112 (S206). Next, the output information generation unit 116 generates output information based on suitable candidates (S208). Next, the output unit 118 outputs the output information to the terminal device 20 or the like (S210). This is the end of the description of the processing of this flowchart.

According to the information processing apparatus 100 of the embodiment described above, the acquisition unit 102 that acquires voice data, the analysis unit 104 that analyzes voice data and converts it into candidate text, and outputs one or more analysis results. The W2V execution unit 106 that converts each of the plurality of words included in the candidate text related to the analysis result into a vector value by distributed expression, the vector value converted by the W2V execution unit 106, and the voice related to the voice data are emitted. An analysis result with a high possibility that the input intention is reflected from one or more analysis results based on the word vector list 120f obtained in advance corresponding to the input text whose input intention of the user's input text is known. By providing the selection unit 114 to be selected, the voice recognition process can be performed more efficiently.

Although the embodiments for carrying out the present invention have been described above using the embodiments, the present invention is not limited to these embodiments, and various modifications and substitutions are made without departing from the gist of the present invention. Can be added.

100 ... Information processing device, 20 ... Terminal device, 30 ... Control target device, 40 ... Service server, 100 ... Information processing device, 102 ... Acquisition unit, 104 ... Analysis unit, 106 ... W2V execution unit, 108 ... Text vector generation unit , 110 ... Sorting unit, 110a ... Reliability derivation unit, 112 ... Language model calculation unit, 114 ... Selection unit, 116 ... Output information generation unit, 118 ... Output unit

Claims (14)

The acquisition unit that acquires audio data,
An analysis unit that analyzes the voice data and converts it into text, and outputs one or more analysis results.
A vector conversion unit that converts each of a plurality of words included in the text related to the analysis result into a vector value by a distributed expression, and a vector conversion unit.
The vector value converted by the vector conversion unit and the input text of the user who has emitted the voice related to the voice data correspond to the input text whose input intention is known, and the vector of the known input text obtained in advance. A selection unit that selects the analysis result that is likely to reflect the input intention from the analysis results of 1 or more based on the value.
A reliability derivation unit for deriving the reliability of the analysis result using the posterior probability of the word obtained as the search result of the speech recognition engine is provided.
The selection unit changes the analysis result to be selected based on the reliability.
Information processing equipment. The selection unit preferentially selects the analysis result whose reliability is equal to or higher than the threshold value.
The information processing apparatus according to claim 1 . The acquisition unit that acquires audio data,
An analysis unit that analyzes the voice data and converts it into text, and outputs one or more analysis results.
A vector conversion unit that converts each of a plurality of words included in the text related to the analysis result into a vector value by a distributed expression, and a vector conversion unit.
The vector value converted by the vector conversion unit and the input text of the user who has emitted the voice related to the voice data correspond to the input text whose input intention is known, and the vector of the known input text obtained in advance. A selection unit that selects the analysis result that is likely to reflect the input intention from the analysis results of 1 or more based on the value.
Equipped with
The vector conversion unit derives a representative vector of a cluster which is a group of the known input texts having a semantic similarity of a predetermined degree or more.
The selection unit selects the first stage of the analysis result using the representative vector, and then reflects the input intention of the user's input text from the cluster selected by the selection of the first stage. Select the analysis result that is likely to be
Information processing equipment. The acquisition unit that acquires audio data,
An analysis unit that analyzes the voice data and converts it into text, and outputs one or more analysis results.
A vector conversion unit that converts each of a plurality of words included in the text related to the analysis result into a vector value by a distributed expression, and a vector conversion unit.
The vector value converted by the vector conversion unit and the input text of the user who has emitted the voice related to the voice data correspond to the input text whose input intention is known, and the vector of the known input text obtained in advance. A selection unit that selects the analysis result that is likely to reflect the input intention from the analysis results of 1 or more based on the value.
Equipped with
The selection unit determines whether or not the voice data includes the user's task execution intention based on the position information given to the voice data.
Information processing equipment. The selection unit changes the accuracy of selection of the corresponding task according to the input environment of the voice data estimated based on the position information.
The information processing apparatus according to claim 4 . The acquisition unit that acquires audio data,
An analysis unit that analyzes the voice data and converts it into text, and outputs one or more analysis results.
A vector conversion unit that converts each of a plurality of words included in the text related to the analysis result into a vector value by a distributed expression, and a vector conversion unit.
The vector value converted by the vector conversion unit and the input text of the user who has emitted the voice related to the voice data correspond to the input text whose input intention is known, and the vector of the known input text obtained in advance. A selection unit that selects the analysis result that is likely to reflect the input intention from the analysis results of 1 or more based on the value.
Equipped with
Further, an output information generation unit for outputting an instruction related to a task for generating output information corresponding to the input intention based on the selection result by the selection unit is further provided.
Information processing equipment. The computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed representation indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance, the input intention of the input text of the user who emitted the voice related to the voice data corresponds to the known input text. , Select the analysis result having a high possibility of reflecting the input intention from the analysis results of 1 or more.
The reliability of the analysis result is derived by using the posterior probability of the word obtained as the search result of the speech recognition engine.
At the time of the selection, the analysis result selected based on the reliability is changed.
Information processing method. The computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed representation indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance corresponding to the input text whose input intention of the input text of the user who emitted the voice related to the voice data is known. , Select the analysis result having a high possibility of reflecting the input intention from the analysis results of 1 or more.
When converting to the vector value, a representative vector of a cluster, which is a group of the known input texts having a semantic similarity of a predetermined degree or more, is derived.
At the time of the selection, the first stage selection of the analysis result is performed using the representative vector, and then the input intention of the input text of the user is reflected from the cluster selected by the selection of the first stage. Select the analysis result that is likely to have been done,
Information processing method. The computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed representation indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance corresponding to the input text whose input intention of the input text of the user who emitted the voice related to the voice data is known. , Select the analysis result having a high possibility of reflecting the input intention from the analysis results of 1 or more.
At the time of the selection, it is determined whether or not the voice data includes the execution intention of the user's task based on the position information given to the voice data.
Information processing method. The computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed representation indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance corresponding to the input text whose input intention of the input text of the user who emitted the voice related to the voice data is known. , Select the analysis result having a high possibility of reflecting the input intention from the analysis results of 1 or more.
Based on the result of the selection, an instruction relating to a task for generating output information corresponding to the input intention is output.
Information processing method. On the computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed expression indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance, the input intention of the input text of the user who emitted the voice related to the voice data corresponds to the known input text. , The analysis result having a high possibility of reflecting the input intention is selected from the analysis results of one or more .
The reliability of the analysis result is derived by using the posterior probability of the word obtained as the search result of the speech recognition engine.
At the time of the selection, the analysis result selected based on the reliability is changed.
program. On the computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed expression indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance, the input intention of the input text of the user who emitted the voice related to the voice data corresponds to the known input text. , The analysis result having a high possibility of reflecting the input intention is selected from the analysis results of one or more.
When converting to the vector value, a representative vector of a cluster, which is a group of the known input texts having a semantic similarity of a predetermined degree or more, is derived.
At the time of the selection, the representative vector is used to select the first stage of the analysis result, and then the input intention of the user's input text from the cluster selected by the selection of the first stage is Select the analysis result that is likely to be reflected,
program. On the computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed expression indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance corresponding to the input text whose input intention of the input text of the user who emitted the voice related to the voice data is known. , The analysis result having a high possibility of reflecting the input intention is selected from the analysis results of one or more.
At the time of the selection, it is determined whether or not the voice data includes the execution intention of the user's task based on the position information given to the voice data.
program. On the computer
Get voice data,
One or more analysis results obtained by analyzing the voice data and converting it into text are output.
It is converted into a vector value by a distributed expression indicating each of a plurality of words included in the text related to the analysis result.
Based on the converted vector value and the vector value of the known input text obtained in advance corresponding to the input text whose input intention of the input text of the user who emitted the voice related to the voice data is known. , The analysis result having a high possibility of reflecting the input intention is selected from the analysis results of one or more.
Based on the result of the selection, an instruction relating to a task for generating output information corresponding to the input intention is output.
program.

Images