JP4110012B2 - Conversation control device and conversation control method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a conversation control device and a conversation control method.
[0002]
[Prior art]
A conventional conversation control device can output a pre-stored answer sentence corresponding to the utterance content based on the utterance content from the user (for example, Patent Document 1). For this reason, the user was able to experience a feeling of talking with another user in a pseudo manner by using the reply sentence from the conversation control device.
[0003]
[Patent Document 1]
JP 2002-169804 (page 10-26, FIG. 16)
[0004]
[Problems to be solved by the invention]
However, when the user's utterance content is captured in a plurality of meanings and there are a plurality of answer sentences corresponding to the utterance contents, the conversation control device outputs the answer sentence searched first. For this reason, when the user's utterance content is captured in a plurality of meanings, the conversation control device captures the meaning of the utterance content related to a specific topic to be developed in advance, and corresponds to the captured utterance content. Answer text could not be output.
[0005]
For example, if the utterance content from the user is {Tell me about IR}, the utterance content is captured in the meaning of IR (Inazumarush) of the production image on the gaming machine or IR (Investor Relations) of the company Sometimes. In this case, when the conversation control device first searches for an answer sentence corresponding to the IR (Inazu mash) of the effect image on the gaming machine from among the answer sentences, the conversation control apparatus outputs the answer sentence searched first. .
[0006]
That is, even when the conversation control device wants to output an answer sentence corresponding to the IR of the company, when it first searches for an answer sentence corresponding to the IR (Inazu mash) of the production image, the answer that is searched first The sentence was output automatically. For this reason, it has been desired to develop a system that can output an answer sentence related to a specific topic to be developed in advance.
[0007]
Therefore, the present invention has been made in view of the above points, and when a user utters a plurality of the same utterance contents captured in a plurality of meanings, an answer sentence different from the previous answer sentence is output. It is an object of the present invention to provide a conversation control device and a conversation control method that can be used.
[0008]
[Means for Solving the Problems]
In order to solve the above problem, the present invention is based on the input information input from the user, and when the same input information is first acquired, it is acquired first. It is characterized in that an answer sentence different from the answer sentence corresponding to the same input information is searched.
[0009]
That is, the present invention searches each phrase including input information from a plurality of phrases stored in advance based on input information input from a user, and searches the previous time based on each searched phrase. The phrase searched is included in each phrase searched this time, and when a predetermined time has passed since the previous phrase was searched, a phrase different from the phrase searched last time is changed to each phrase searched this time. And selecting an answer sentence corresponding to the phrase based on the selected phrase.
[0010]
In this case, for example, phrase 1 (tell IR of device A) and phrase 2 (tell IR of device B) are searched as phrases including the searched input information (tell IR). The phrase 1 (tell me the IR of the device A) was included in each of these phrases, and a predetermined time from when the phrase 1 (tell me the IR of the device A) was last searched. When the time has elapsed, the conversation control device can select a phrase 2 different from the previously searched phrase 1 and output an answer sentence corresponding to the selected phrase 2.
[0011]
As a result, when the input information is captured in a plurality of meanings, the meanings used last time are included in each meaning, and the predetermined time has elapsed since the meaning was used, the conversation control device Depending on the time, a meaning different from the meaning used last time can be selected, and an answer sentence corresponding to the selected meaning can be output. As a result, when the user utters a specific utterance content and the same content as the utterance content is repeatedly input after a predetermined time has elapsed since the utterance, the conversation control device An answer sentence that directs the current topic to a different topic according to the time can be output.
[0012]
In the said invention, the phrase contains the 2nd morpheme information which shows the morpheme which consists of one character, several character strings, or these combination, and the response sentence to a user is matched with several 2nd morpheme information. Each of the second morpheme information and each answer sentence is stored in advance, and at least one morpheme constituting the minimum unit of the character string is determined based on the character string included in the input information input by the user. Based on the extracted first morpheme information, a plurality of second morpheme information including the first morpheme information is searched from the previously stored second morpheme information, and the second morpheme searched last time If the information is included in each second morpheme information searched this time, and a predetermined time has elapsed since the previous second morpheme information was searched, the second morpheme information different from the previously searched second morpheme information Morphology And elected from among the second morpheme information retrieved this time, based on the second morpheme information elected, it is preferable to find the answer sentence associated with the second morpheme information.
[0013]
According to the invention according to the present application, the second morpheme information searched last time is included in each second morpheme information searched this time, and a predetermined time has elapsed since the last time the second morpheme information was searched. In this case, the conversation control device selects second morpheme information different from the previously searched second morpheme information from each second morpheme information searched this time, and corresponds to the selected second morpheme information. Attached answer sentences can be output.
[0014]
Specifically, for example, when a user utters “Tell me financial information” {First morpheme information (financial information; tell me)}, the contents of the utterance include Company A, Company B・ Financial information on multiple companies is included. Here, the fact that a plurality of second morpheme information including the first morpheme information constituting the utterance content of the user is searched means that the utterance content of the user has an ambiguous meaning.
[0015]
In this case, the conversation control device, for example, as the second morpheme information including the first morpheme information (financial information; tell me), (company A; financial information; tell me), (company B; financial information; tell me) ) ... will be searched. At this time, if the conversation control device searches the second morpheme information (A company; financial information; tell me) last time and a predetermined time has passed since the previous search, the conversation control device Second morpheme information (company B; financial information; tell me) different from morpheme information (company A; financial information; tell me) is searched. The conversation control device outputs an answer sentence (the financial information of company B is ~) associated with the searched second morpheme information (company B; financial information; tell me).
[0016]
That is, when the user utters a plurality of the same utterance contents that can be captured in a plurality of meanings and a predetermined time has elapsed since the utterance was first made, the conversation control device is different from the previous answer sentence. Output answer text. As a result, if the predetermined time is set to be short, the answer sentences are changed one by one within a short time, so that the conversation control device gives the user an impression that it is easy to forget the contents that have been answered before. It can be used as a device that can.
[0017]
On the other hand, if the predetermined time is set to be long, the answer sentence is not changed so much, so the conversation control device is a device that can give the user the impression of persistently speaking the content that was previously answered. Can be used.
[0018]
In particular, the conversation control device searches for a corresponding answer sentence using a morpheme excluding Japanese-specific particles included in the input information, and grasps it from each morpheme. Based on the meaning content that can be made, it is possible to search for a reply sentence prepared in advance more accurately. That is, the conversation control device can search for an answer sentence more suitable for the input information than simply searching for an answer sentence associated with the keyword using the entire input information as a keyword.
[0019]
Furthermore, since the conversation control device searches for the second morpheme information including the first morpheme information, it is not necessary to search for the second morpheme information that completely matches the input information from the user. As a result, the developer who develops the conversation control device does not need to store a huge amount of answer sentences corresponding to input information expected to be input from the user in advance, and can reduce the capacity of the storage unit. .
In addition, it is preferable that the answer sentence different from the previous answer sentence includes the content to be developed in advance. In this case, for example, when the same utterance content is repeatedly uttered, the conversation control device can output an answer sentence that directs the current topic to another topic.
[0020]
As a result, when the same utterance content as the previous time is input and a predetermined time has elapsed since the input of the same utterance content, for example, the character appearing in the game changes the current story to another story. An answer sentence for changing can be output.
[0021]
In the above invention, a plurality of second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with an answer sentence to the user, and each second morpheme Information and each answer sentence are stored in advance, and at least one morpheme constituting the minimum unit of the character string is extracted as the first morpheme information and extracted based on the character string included in the input information input from the user. Based on the first morpheme information, a plurality of second morpheme information including the first morpheme information is searched from each second morpheme information stored in advance, and the previously searched second morpheme information is searched this time. If each second morpheme information that is included in each second morpheme information and searched this time exceeds a predetermined reference number of times, second morpheme information that is different from the previously searched second morpheme information, Searched this time Elected from among the second morpheme information based on the second morpheme information elected, it is preferable to find the answer sentence associated with the second morpheme information.
[0022]
In this case, the second morpheme information searched last time is included in each second morpheme information searched this time, and each second morpheme information searched this time is searched beyond a predetermined reference number. The conversation control device selects second morpheme information different from the previously searched second morpheme information from the second morpheme information searched this time, and a response associated with the selected second morpheme information. A sentence can be output.
[0023]
In other words, if the user utters the same utterance content captured in multiple meanings and the same utterance content is uttered beyond the reference count, the conversation control device is different from the previous answer sentence. Output answer text. As a result, if the reference number is set to be small, the answer sentences are changed one by one, so that the conversation control device can give the user the impression that it is easy to forget the contents that were answered before. Can be used as
[0024]
On the other hand, if the reference number is set to a large number, the answer text is not changed so much, so the conversation control device is a device that can give the user the impression of persistently speaking the content that was answered before. Can be used.
[0025]
DETAILED DESCRIPTION OF THE INVENTION
[First embodiment]
(Basic configuration of conversation control system)
A conversation control system according to the present invention will be described with reference to the drawings. FIG. 1 is a schematic configuration diagram of a conversation control system having a conversation control apparatus 1 according to the present embodiment.
[0026]
As shown in the figure, the conversation control device 1 includes an input unit 100, a speech recognition unit 200, a conversation control unit 300, a sentence analysis unit 400, a conversation database 500, an output unit 600, and a speech recognition dictionary storage. Part 700.
[0027]
In the present embodiment, for convenience of explanation, the description is limited to the user's utterance content (this utterance content is a kind of input information), but is not limited to the user's utterance content, and the keyboard. The input information may be input from the above. Therefore, the “utterance content” shown below can be described by replacing “utterance content” with “input information”.
[0028]
Similarly, in the following description, for convenience of explanation, the description will be limited to the “spoken sentence type” (speech type), but is not limited to this “spoken sentence type”, and input from a keyboard or the like. It may be an “input type” indicating the type of input information. Accordingly, the following “speech sentence type” (speech type) can be described by replacing “speech type” with “input type”.
[0029]
The input unit 100 is an acquisition unit that acquires input information from a user. In the present embodiment, a microphone, a keyboard, and the like are used. The input unit 100 is also a character recognition unit that specifies a character string indicating input information based on input information input from a user.
[0030]
Here, the input information means characters, symbols, voices and the like input through a keyboard or the like. Specifically, the input unit 100 acquires user input information (other than voice), and outputs the acquired input information to the conversation control unit 300. Further, the input unit 100 that has acquired the utterance content from the user (the utterance content is composed of voice and is a kind of input information) with a microphone or the like, the voice constituting the acquired utterance content is a voice signal. To the voice recognition unit 200.
[0031]
The voice recognition unit 200 is a character recognition unit that identifies a character string included in the utterance content based on the utterance content acquired by the input unit 100. Specifically, the speech recognition unit 200 to which a speech signal is input from the input unit 100 analyzes the input speech signal, and a character string corresponding to the analyzed speech signal is stored in the speech recognition dictionary storage unit 700. The specified character string is output to the conversation control unit 300 as a character string signal. The speech recognition dictionary storage unit 700 stores a dictionary corresponding to standard speech signals.
[0032]
The sentence analysis unit 400 analyzes a character string specified by the input unit 100 or the speech recognition unit 200. In the present embodiment, as shown in FIG. 2, a morpheme extraction unit 410 and a phrase analysis unit 420 are included. A sentence structure analysis unit 430, an utterance type determination unit 440, a morpheme database 450, and an utterance type database 460.
[0033]
The morpheme extraction unit 410 is a morpheme extraction unit that extracts, as first morpheme information, at least one morpheme constituting a minimum unit of a character string based on the character string specified by the input unit 100 or the speech recognition unit 200.
[0034]
Specifically, the morpheme extraction unit 410 to which the character string signal is input from the management unit 310 extracts each morpheme from the character string corresponding to the input character string signal. Here, in this embodiment, the morpheme means the minimum unit of the word structure represented in the character string. As the minimum unit of the word structure, as shown in FIG. 3, for example, parts of speech such as nouns, adjectives and verbs can be cited. In the present embodiment, each morpheme is expressed as m1, m2,.
[0035]
That is, the morpheme extraction unit 410 collates a character string corresponding to the input character string signal with a morpheme group such as nouns, adjectives, and verbs stored in advance in the morpheme database 450, and the morpheme is extracted from the character string. Each morpheme (m1, m2,...) That matches the group is extracted, and each extracted morpheme is output to the phrase analysis unit 420 as an extraction signal.
[0036]
The phrase analysis unit 420 is a conversion unit that converts each morpheme into a phrase format based on each morpheme extracted by the morpheme extraction unit 410. Specifically, the phrase analysis unit 420 to which the extraction signal is input from the morpheme extraction unit 410 uses the morphemes corresponding to the input extraction signal to combine them into a phrase format.
[0037]
Here, in this embodiment, the phrase format is a sentence in which the independent grammar or one or more attached words are attached to the independent grammar in the Japanese grammar, or a character string that does not destroy the meaning of the Japanese grammar. Means a sentence that is separated as finely as possible. This clause is expressed as p1, p2,... Pk in this embodiment.
[0038]
That is, the phrase analysis unit 420, as shown in FIG. 4, based on each morpheme corresponding to the input extracted signal, the dependency element of each morpheme (for example, (m2), (m4), (m5) ) ・ ・） Is extracted, and each morpheme is grouped into each phrase based on the extracted dependency elements. “T” shown in the figure means transposition.
[0039]
The phrase analysis unit 420 that collects each morpheme into each phrase includes a sentence structure analysis unit 430 and an utterance type determination unit 440 using sentence pattern information including each phrase that combines each morpheme and each morpheme constituting each phrase as a sentence pattern signal. Output to.
[0040]
The sentence structure analysis unit 430 is a classification unit that classifies each morpheme of the first morpheme information segmented by the phrase analysis unit 420 into attributes such as subject case and target case. Specifically, the sentence structure analysis unit 430, to which the sentence pattern signal is input from the phrase analysis unit 420, determines each morpheme included in the phrase based on each morpheme corresponding to the input sentence pattern signal and the phrase composed of each morpheme. Determine the “case composition”.
[0041]
Here, the “case structure” means a case (attribute) indicating a substantial concept in the clause. In the present embodiment, for example, a subject (subject) that represents a subject / subject, Examples include a meaning object (target case), an action meaning an action, a time meaning (tense, aspect), a location meaning a place, and the like. In this embodiment, each morpheme associated with the “case” (case configuration) of the three elements of the subject, the object, and the action is used as the first morpheme information.
[0042]
That is, as shown in FIG. 5, for example, when the dependency element of each morpheme is “” or “is”, the sentence structure analysis unit 430 applies the subject (subject) to the morpheme before the dependency element. Or a leading character). For example, when the dependency element of each morpheme is “NO” or “NO”, the sentence structure analysis unit 430 determines that the morpheme before the dependency element is an object (target).
[0043]
Further, for example, when the dependency element of each morpheme is “Yes”, the sentence structure analysis unit 430 includes an action (predicate; this predicate is composed of a verb, an adjective, and the like). It is determined that
[0044]
The sentence structure analysis unit 430 that has determined the “case structure” of each morpheme that constitutes each clause determines the range of topics (topics) to be described later based on the first morpheme information associated with the determined “case structure”. A topic search command signal for specifying is output to the reflective determination unit 320.
[0045]
The utterance type determination unit 440 is a type specifying unit that specifies an utterance type (input type) indicating the type of utterance content (input information) based on the phrase specified by the phrase analysis unit 420. Specifically, “speech sentence type” (speech type) is determined based on each morpheme corresponding to the sentence pattern signal input from the phrase analysis unit 420 and a phrase composed of each morpheme.
[0046]
In this embodiment, as shown in FIG. 6, the “spoken sentence type” is a statement sentence (D; Declaration), an impression sentence (I; Impression), a conditional sentence (C; Condition), and a result sentence ( E; Effect, time sentence (T; Time), location sentence (L; Location), repulsive sentence (N; Negation), and the like. Each of these sentences is expressed by an affirmative sentence (A; Answer) or a question sentence (Q; Question) as shown in FIG.
[0047]
The statement sentence means a sentence composed of a user's opinion or idea, and in this embodiment, as shown in FIG. 6, for example, a sentence such as “I like Sato” can be cited. An impression sentence means the sentence which consists of an impression which a user holds. A place sentence means a sentence made up of place elements.
[0048]
The result sentence means a sentence composed of sentences including a result element for a topic. A time sentence means a sentence composed of sentences including temporal elements related to a topic.
[0049]
The conditional sentence means a sentence composed of sentences including elements such as a premise of a topic, a condition and a reason why the topic is established, when one utterance is regarded as a topic. The repulsive sentence means a sentence composed of a sentence including an element that repels the utterance partner. An example sentence for each “spoken sentence type” is as shown in FIG.
[0050]
That is, the utterance type determination unit 440 collates each phrase with each dictionary stored in the utterance type database 460 based on each phrase corresponding to the input sentence pattern signal, and from each phrase, Extract sentence elements related to the dictionary. The utterance type determination unit 440 that extracts sentence elements related to each dictionary from each phrase determines “spoken sentence type” based on the extracted sentence elements. The sentence element means a type for identifying the type of character string, and the sentence element in the present embodiment includes the above-described definition phrase (which means “-”).
[0051]
Here, as shown in FIG. 7, the utterance type database 460 includes a definition expression example dictionary having a dictionary related to a definition phrase (for example,), an affirmative phrase (for example, approval, sympathy, ping pong). positive cases dictionary with a dictionary related to, a result clause (for example, so, so) the results representation case dictionary with a dictionary related to, greeting clause (for example, Hello) greeting case dictionary with a dictionary related to, It is composed of a negative example dictionary having a dictionary related to a negative phrase (for example, it is not idiot, the opposite), and each dictionary is associated with “type of spoken sentence”.
[0052]
As a result, the utterance type determination unit 440 compares the phrase with each dictionary stored in the utterance type database 460, extracts the sentence element related to each dictionary from the phrase, and associates it with the extracted sentence element. The “spoken sentence type” can be determined by referring to the determined type.
[0053]
The utterance type determination unit 440 outputs an answer search command signal for causing the corresponding user to search for a specific answer sentence to the answer sentence search unit 370 based on an instruction from the topic search unit 360 described later.
[0054]
The conversation database 500 stores in advance a plurality of second morpheme information indicating each morpheme made up of one character, a plurality of character strings, or a combination thereof, and a reply sentence to the user for the utterance content in association with each other. It is a lottery storage means. The conversation database 500 also stores a plurality of answer types indicating the types of answer sentences associated with a plurality of answer sentences in advance in association with the second morpheme information.
[0055]
Furthermore, the conversation database 500 stores in advance a plurality of discourse ranges (keywords) indicating morphemes that constitute a range relevant to the input content that will be input by the user or the answer sentence to the user. . A plurality of second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with the discourse range (keyword), and each second morpheme information includes an answer sentence to the user. Associated with each.
[0056]
Furthermore, the conversation database 500 classifies and stores each element constituting the second morpheme information as an attribute such as each subject consisting of a main case and a target case consisting of a target case.
[0057]
As shown in FIG. 8, in the present embodiment, the conversation database 500 is roughly divided in the present embodiment, which means a range of relevance with respect to utterance contents that will be uttered by the user or an answer sentence to the user. It consists of a range (discourse) and a topic (topic) that means a range that is most closely related to the content the user is speaking. As shown in the figure, “discourse range” is positioned as a superordinate concept of “topic” in the present embodiment.
[0058]
Each discourse range can be configured to have a hierarchical structure as shown in FIG. As shown in the figure, for example, a higher level discourse range (entertainment) for a certain discourse range (movie) is positioned in the upper hierarchical structure, and a lower level discourse range (movie) for the discourse range (movie). The attribute (movie) can be located in the lower hierarchical structure. That is, in the present embodiment, each discourse range can be arranged at a hierarchical position where the relationship between the higher concept, the lower concept, the synonym, and the synonym becomes clear with other discourse ranges.
[0059]
As described above, the discourse range is composed of topics, and in this embodiment, for example, if the discourse range is an A movie name, it includes a plurality of topics related to the A movie name.
[0060]
This topic means a morpheme composed of a single character, a plurality of character strings, or a combination thereof, that is, each morpheme constituting speech content that will be uttered by the user. Each morpheme is associated with a subject (subject), an object (target case), and an action “case” (attribute). In this embodiment, each morpheme associated with these three elements is expressed as a topic title (this topic title corresponds to a subordinate concept of “topic”) (second morpheme information). .
[0061]
The topic title is not limited to each morpheme associated with the above three elements, but other “cases”, that is, time (tense, aspect) meaning time, location meaning place, You may have each morpheme matched with the condition which means a condition, the impression which means an impression, the effect which means a result, etc.
[0062]
In this embodiment, the topic title (second morpheme information) is stored in advance in the conversation database 500, and the first morpheme information (derived from the utterance content uttered by the user) is It is a distinction.
[0063]
For example, if the talk range is “A movie name”, as shown in FIG. 10, the subject title is subject (A movie name), object (director), action (great) {this is “A movie name” The director of the name is composed of "meaning great".
[0064]
If there is no morpheme associated with “case composition” (subject, object, action, etc.) among the topic titles, “*” is indicated for the portion in the present embodiment.
[0065]
For example, {A movie name? } Is converted into a topic title (subject; object; action). }, “A movie name” can be specified as a subject, but “object” and “action” are not elements of the sentence, so the topic title is “subject” (A movie name) “No object” (*); no “action” (*) (see FIG. 10).
[0066]
In the present embodiment, the answer sentence is associated with each topic title (second morpheme information) (see FIG. 8). In this embodiment, as shown in FIG. 11, the answer sentence is a statement sentence (D; Declaration), an impression sentence in order to make an answer corresponding to the type of utterance sentence uttered by the user. Types such as (I; Impression), conditional statement (C; Condition), result statement (E; Effect), time statement (T; Time), location statement (L; Location), negative statement (N; Negation) Type).
[0067]
That is, as shown in FIG. 12, each answer sentence is associated with, for example, a discourse range (Sato) {subordinate concept; home run, superordinate concept; grass baseball, synonym; panda Sato, Sato player, panda} and each topic title. It has been.
[0068]
As shown in the figure, for example, the topic title 1-1 is {(Sato; *; I like): this consists of the order of (subject; object; action) as described above. If the order is the same below, the answer sentence 1-1 corresponding to the topic title 1-1 is (DA; statement affirmation “I like Sato”), (IA; comment affirmation) "I like Sato very much"), (CA; conditional affirmation sentence "Sato's home run is very impressive"), (EA; a result affirming sentence "I always watch Sato's games on TV"), (TA: Time affirmative sentence "I actually like it from the five-bats continual refrain in Koshien"), (LA; Place affirmative sentence "I like the serious face when standing on the blow"), (NA A repulsive affirmative sentence "I don't want to talk to people who don't like Sato, goodbye").
[0069]
In the present embodiment, as shown in FIG. 2, the conversation control unit 300 includes a management unit 310, a reflexive determination unit 320, a turnback determination unit 330, a discourse range determination unit 340, and an abbreviated sentence complement unit 350. , A topic search unit 360 and an answer sentence search unit 370.
[0070]
The management unit 310 controls the entire conversation control unit 300. Specifically, the management unit 310 to which a character string is input from the input unit 100 or the voice recognition unit 200 outputs the input character string to the morpheme extraction unit 410 as a character string signal. In addition, the management unit 310 outputs the answer sentence searched by the answer sentence search unit 370 to the output unit 600.
[0071]
The reflexive determination unit 320 is a fixed form acquisition unit that collates the first morpheme information extracted by the morpheme extraction unit 410 with each fixed content, and searches the fixed content including the first morpheme information from each fixed content. .
[0072]
Here, the standard content means reflection element information for replying a standard content to the utterance content from the user, and this reflection element information is stored in the reflection element database 801 (standard storage means) in advance. A plurality are stored. As the reflective element information, in the present embodiment, as shown in FIG. 13, for example, "Good morning", "Hello", "Good evening", "greeting elements" such as "Hey", "I see", "Really?""Typicalelements" such as
[0073]
Specifically, the reflexive determination unit 320 to which the topic search command signal is input from the sentence structure analysis unit 430 receives the first morpheme information included in the input topic search command signal and the reflection element database 801. The reflection element information is collated, the reflection element information including the first morpheme information is searched from the reflection element information, and the searched reflection element information is output to the management unit 310.
[0074]
That is, assuming that the reflection element information is D1 and the first morpheme information is W, the reflective determination unit 320 determines that the relationship of W∩D1 ≠ φ (φ; empty set) is established. Process to give a typical answer.
[0075]
For example, when the user utters the utterance content “Good morning”, the reflexive determination unit 320 collates the utterance content “Good morning” with each reflection element information, and the utterance content from each reflection element information. The reflection element information “good morning” including (matching) “good morning” is searched, and the searched reflection element information “good morning” is output to the management unit 310.
[0076]
When the reflection determination unit 320 cannot search the reflection element information including the utterance content from the reflection element information, the reflex determination unit 320 returns the topic search command signal input from the sentence structure analysis unit 430. Output to.
[0077]
The return determination unit 330 collates the current first morpheme information extracted by the morpheme extraction unit 410 with a past answer sentence stored in the return element database 802, and the current first morpheme information is a past answer. When included in the sentence, it is a fixed form acquisition means for acquiring the contents of the agreement.
[0078]
Here, “turnback” means to say back the content of the user's utterance as it is (or content close to it) in this embodiment. In this embodiment, the return element is composed of first morpheme information constituting the answer sentence output from the conversation control apparatus 1 immediately before, for example, as shown in FIG. (Horse; *; Beautiful), “I like Sato” (Sato; *; I like).
[0079]
Further, the return element database 802 is also an agreement storage means for preliminarily storing agreement contents for agreeing to input information input by the user. The content of the agreement includes, for example, the input information previously input by the user (if the previous input information by the user is “A film name director is Mr. S”, the content of the agreement is “ The director of the A movie name is Mr. S)), or “Yes”, “It is true”.
[0080]
Specifically, the return determination unit 330 to which the topic search command signal is input from the reflective determination unit 320 configures the first morpheme information and the return element included in the input topic search command signal for each return element. Each morpheme is collated, and it is determined whether or not the first morpheme information is included in the return element (see FIG. 14).
[0081]
If it is determined that the first morpheme information is included in each return element, the return determination unit 330 acquires the content of the agreement and outputs an answer sentence including the acquired content of the agreement to the management unit 310 (return) Process). That is, assuming that the return element (such as the previous answer sentence) is S and the first morpheme information is W, the return determination unit 330 indicates the above when the relationship of W⊂S and W ≠ φ is established. Performs a reversal process.
[0082]
For example, the conversation control device 1 responds with “The director of the A movie name is Mr. S” (Director of the A movie name; Mr. S; *) (the order is the subject, the object, the order of the action, and so on. ) Is output, and if the user utters “Is the director of the movie name Mr. S” (Director of the movie name; Mr. S; Since the determination unit 330 matches the first morpheme information of the user (Director of A movie name; Mr. S; *) and each morpheme of the answer sentence (Director of A movie name; Mr. S; *), The user concludes that he / she is replying to the reply sentence, obtains the stored agreement contents “Yes”, and outputs the obtained agreement contents.
[0083]
Further, the return determination unit 330 collates the current first morpheme information extracted by the morpheme extraction unit 410 with the past first morpheme information stored in the return element database 802 to obtain the current first morpheme information. Is included in the past first morpheme information, it is also a fixed form acquisition means for acquiring the repulsion content.
[0084]
Specifically, when the user utters the utterance content “Horse is beautiful” and the conversation control device 1 outputs the content “Horse is good” as an answer sentence, it is used later When the person repeats the utterance content that “the horse is beautiful”, the turn-back determination unit 330 determines that each morpheme (first morpheme information) {horse; *; beautiful} constituting the current utterance content “horse is beautiful” Each morpheme (first morpheme information) {horse; *; beautiful} that constitutes the content of the previous utterance “Horse is beautiful” is the same, so the user responds from the conversation control device 1 “Horse is dynamic You can conclude that you have not heard about "I feel good."
[0085]
In this case, since the user has not heard the answer sentence from the conversation control device 1, the return determination unit 330 acquires the stored repulsion content (for example, do not repeat the same content) and the acquired repulsion The contents can be output.
[0086]
On the other hand, if the reversion determining unit 330 determines that the first morpheme information is the same as the content of the previous answer sentence or the first morpheme information is not the same as the previous first morpheme information, the reflex determination unit 320 The input topic search command signal is output to the conversation range determination unit 340.
[0087]
In addition, although the said return determination part 330 has shown the process in case a user makes a return with respect to "the reply sentence of the conversation control apparatus 1," the following processes can also be performed. For example, when the output unit 600 outputs an answer sentence “Horse is beautiful”, the user answers “Why is the horse beautiful?”, “Why is it beautiful?”, Or “Why?” This is a process of the turn-back determination unit 330 performed when the utterance is made.
[0088]
In this case, the return determination unit 330 outputs the response sentence S “the horse is beautiful” and the utterance content W (“why is the horse beautiful? (Question sentence)” or “why is it beautiful? (Question sentence) ”), (W−c) ⊂S, S ≠ φ, c ≠ φ (where c means the utterance type of W, and this utterance type is determined by the utterance type determination unit 440 described later. As will be described later, for example, a question sentence can be cited as an utterance type, so a “conditional” reversal process (the user asks the question sentence for the answer sentence). (Processing when a flip is performed).
[0089]
For example, when the conversation control device 1 outputs an answer sentence “The horse is beautiful”, the above-mentioned user utters the content of the utterance “Why is the horse beautiful?”. At that time, in order to resolve the user's question, the return determination unit 330 acquires an answer sentence such as “Even the horse is not beautiful” from the return element database 802, and the acquired answer sentence is sent to the management unit 310. Process to output.
[0090]
The discourse range determination unit 340 collates the first morpheme information extracted by the phrase analysis unit 420 with each discourse range stored in advance, and matches the morpheme included in the first morpheme information from each discourse range. This is a discourse range search means for searching a discourse range.
[0091]
Specifically, the conversation range determination unit 340 to which the topic search command signal is input from the turn-back determination unit 330 determines the user's conversation range based on the input talk search command signal. That is, based on the input search command signal, the conversation range determination unit 340 searches the conversation database 500 for a range (discourse range) relevant to the content spoken by the user.
[0092]
For example, the discourse range determination unit 340 includes the first morpheme information included in the input topic search command signal (an interesting movie; *; is there) {is there an interesting movie? }, The first morpheme information is compared with the discourse range group, and if the discourse range group includes a morpheme (for example, “movie”) constituting the first morpheme information, The “movie” included in the information is determined as the discourse range. In this case, since the discourse range “movie” is included in the first morpheme information, the discourse range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the topic search unit 360.
[0093]
On the other hand, when the first morpheme information does not include a discourse range group, the conversation range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the abbreviated sentence complement unit 350.
[0094]
Thereby, the topic search unit 360 described later collates each “topic title” belonging to the “discourse range” determined by the conversation range determination unit 340 with the first morpheme information specified by the sentence structure analysis unit 430. Therefore, it is unnecessary to collate “all” “topic titles” (second morpheme information) with the first morpheme information, and the answer sentence search unit 370 to be described later will search for the final answer sentence. Can be shortened.
[0095]
As described above, the discourse range determination unit 340 collates the first morpheme information and the discourse range group, and if the discourse range group includes the morpheme of the first morpheme information, determines the morpheme as the discourse range. However, the present invention is not limited to this, and the morpheme of the return element retrieved immediately by the return determination unit 330 or the morpheme constituting the utterance content uttered by the user may be determined as the discourse range. . The abbreviated sentence complementing unit 350, which will be described later, can add the discourse range to the first morpheme information from which the morpheme is omitted, using the discourse range determined by the discourse range determination unit 340.
[0096]
The abbreviated sentence complementing unit 350 searches the attributes (subjects, objects, actions, etc.) constituting the first morpheme information based on the first morpheme information extracted by the phrase analysis unit 420 for attributes that do not include a morpheme. Attribute search means. The abbreviated sentence complementing unit 350 is also a morpheme adding unit that adds a morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute based on the searched attribute.
[0097]
Specifically, the abbreviated sentence complementing unit 350 to which the topic search command signal is input from the discourse range determination unit 340 is based on the first morpheme information included in the input discourse search command signal, and the utterance including the first morpheme information. It is determined whether or not the content is an abbreviated sentence. If the utterance content including the first morpheme information is an abbreviated sentence, the morpheme in the discourse range to which the first morpheme information belongs is added to the first morpheme information.
[0098]
For example, the abbreviation sentence complementing unit 350 has a morpheme constituting the first morpheme information included in the input topic search command signal (director; *; *) (director?) (This sentence is “what” director If it is unknown, it means an abbreviated sentence.), The discourse range determined by the discourse range determination unit 340 (A movie name; this A movie name indicates the title of the movie) If it is the first morpheme information belonging to A), the determined discourse range (A movie name) is added to the first morpheme information to the morpheme constituting the first morpheme information (director of “A movie name”; *; *) Yes.
[0099]
That is, assuming that the first morpheme information is W and the determined discourse range is D, the abbreviated sentence complementing unit 350 adds the discourse range D to W and includes the added first morpheme information in the topic search command signal. The information is output to the topic search unit 360.
[0100]
Thereby, even if the first morpheme information is an abbreviated sentence and it is not clear as Japanese, the abbreviated sentence complementing unit 350, when the first morpheme information belongs to a certain discourse range, for example, The discourse range D (A movie name) is added to the first morpheme information W (director; *; *), and the first morpheme information is added to W ′ (A movie name director; *; *). ? }, Even if the user's utterance content is an abbreviated sentence, the abbreviated sentence can be supplemented based on the previously determined discourse range, and the abbreviated sentence should be clarified Can do.
[0101]
For this reason, the first morpheme information is set so that the abbreviated sentence complementing unit 350 has the proper utterance contents constituting the first morpheme information even if the utterance contents constituting the first morpheme information are omitted sentences. Since the specific morpheme can be complemented, the topic search unit 360 acquires the optimal “topic title” (second morpheme information) related to the first morpheme information based on the first morpheme information after the complement. The answer sentence search unit 370 can output an answer sentence that is more suitable for the content of the user's utterance based on the “topic title” acquired by the topic search unit 360.
[0102]
The topic search unit 360 collates the first morpheme information extracted by the phrase analysis unit 420 or the first morpheme information supplemented by the abbreviated sentence complement unit 350 with each second morpheme information, It is the 1st search means which searches the 2nd morpheme information containing the morpheme which comprises 1st morpheme information from inside.
[0103]
Specifically, the topic search unit 360 to which the topic search command signal is input from the talk range determination unit 340 or the abbreviated sentence complement unit 350 is based on the first morpheme information included in the input topic search command signal. The “topic title” including the morpheme of the first morpheme information is searched from each “topic title” (second morpheme information) belonging to the discourse range determined by the determination unit 340, and the search result is used as a search result signal. The information is output to the answer sentence search unit 370 and the utterance type determination unit 440.
[0104]
For example, when the “case configuration” constituting the first morpheme information is (Sato; *; I like) {I like Sato}, the topic search unit 360, as shown in FIG. Each morpheme (Sato; *; I like) belonging to the “composition” is compared with each topic title 1-1 to 1-4 belonging to the discourse range (Sato), and from among the topic titles 1-1 to 1-4 A topic title 1-1 (Sato; *; likes) that matches (or approximates) each morpheme (Sato; *; likes) belonging to the “case composition” is searched, and this search result is used as a search result signal as a response sentence. The data is output to the search unit 370 and the utterance type determination unit 440.
[0105]
The utterance type determination unit 440 to which the search result signal is input from the topic search unit 360, based on the input search result signal, an answer search command for searching for a specific answer sentence that answers to the corresponding user. A signal (this answer search command signal also includes the determined “sentence type”) is output to the answer sentence search unit 370.
[0106]
The answer sentence search unit 370 is an answer acquisition unit that acquires an answer sentence associated with the second morpheme information based on the second morpheme information (topic title) searched by the topic search unit 360. In addition, the answer sentence search unit 370 collates the utterance type of the identified user with each answer type associated with the second morpheme information based on the second morpheme information searched by the topic search unit 360, It is also a second search means for searching for an answer type that matches the user's utterance type from among the answer types.
[0107]
Specifically, the answer sentence search unit 370, to which the search result signal from the topic search unit 360 and the answer search command signal from the utterance type determination unit 440 are input, the topic title (search result) corresponding to the input search result signal. Based on the second morpheme information) and the “speech sentence type” (speech type) corresponding to the answer search command signal, from the group of answer sentences associated with the “topic title”, An answer sentence having an answer type (this answer type means “type of answer sentence” shown in FIG. 11) matching the “sentence type” (DA, IA, CA, etc.) is searched.
[0108]
For example, when the topic title corresponding to the search result is the topic title 1-1 (Sato; *; I like) shown in FIG. 12, the answer sentence search unit 370 is associated with the topic title 1-1. Of the reply sentences 1-1 (DA, IA, CA, etc.), the reply composed of the reply type (DA) that matches the “spoken sentence type” (for example, DA; utterance type) determined by the utterance type determination unit 440. The sentence 1-1 (DA; (I also like Sato) is searched), and the searched answer sentence is output to the management unit 310 as an answer sentence signal.
[0109]
The management unit 310 to which the answer sentence signal is input from the answer sentence search unit 370 outputs the input answer sentence signal to the output unit 600. In addition, the management unit 310 to which the reflection element information is input from the reflective determination unit 320 or the contents of the turn-back process from the turn-back determination unit 330, the answer sentence corresponding to the input reflection element information, the content of the turn-back process that is input To the output unit 600.
[0110]
The output unit 600 is an output unit that outputs the answer text acquired by the answer text search unit 370. In this embodiment, for example, a speaker, a display, and the like can be given. Specifically, the output unit 600 to which an answer sentence is input from the management unit 310 outputs the input answer sentence {for example, I also like Sato}.
[0111]
(Conversation control method using conversation control device)
The conversation control method by the conversation control apparatus 1 having the above configuration can be implemented by the following procedure. FIG. 15 is a flowchart showing the procedure of the conversation control method according to the present embodiment.
[0112]
First, the input unit 100 performs a step of acquiring the utterance content from the user (S101). Specifically, the input unit 100 acquires the voice that constitutes the utterance content of the user, and outputs the acquired voice to the voice recognition unit 200 as a voice signal. Further, the input unit 100 specifies a character string included in the input information (other than the voice) based on the input information (other than the voice) input from the user, and the conversation control unit uses the specified character string as a character string signal. Output to 300.
[0113]
Next, the voice recognition unit 200 performs a step of identifying a character string included in the utterance content based on the utterance content acquired by the input unit 100 (S102). Specifically, the speech recognition unit 200 to which a speech signal is input from the input unit 100 analyzes the input speech signal, and a character string corresponding to the analyzed speech signal is stored in the speech recognition dictionary storage unit 700. The specified character string is output to the conversation control unit 300 as a character string signal.
[0114]
And the morpheme extraction part 410 performs the step which extracts each morpheme which comprises the minimum unit of a character string as 1st morpheme information based on the character string specified by the speech recognition part 200 (S103).
[0115]
Specifically, the morpheme extraction unit 410 to which the character string signal is input from the management unit 310 includes a character string corresponding to the input character string signal and nouns, adjectives, verbs, and the like stored in advance in the morpheme database 450. The morpheme group is collated, each morpheme (m1, m2,...) Matching the morpheme group is extracted from the character string, and each extracted morpheme is output to the phrase analysis unit 420 as an extraction signal.
[0116]
Then, the phrase analysis unit 420 performs a step of grouping each morpheme into a phrase format based on each morpheme extracted by the morpheme extraction unit 410 (S104). Specifically, the phrase analysis unit 420 to which the extraction signal is input from the morpheme extraction unit 410 uses the morphemes corresponding to the input extraction signal to combine them into a phrase format.
[0117]
That is, as shown in FIG. 4, the phrase analysis unit 420 extracts and extracts the dependency elements (for example, gahahahahaha ...) of each morpheme based on each morpheme corresponding to the input extraction signal. Each morpheme is grouped into each clause based on the dependency element.
[0118]
The phrase analysis unit 420 that collects each morpheme into each phrase includes a sentence structure analysis unit 430 and an utterance type determination unit 440 using sentence pattern information including each phrase that combines each morpheme and each morpheme constituting each phrase as a sentence pattern signal. Output to.
[0119]
Thereafter, the sentence structure analysis unit 430 performs a step of classifying each morpheme of the first morpheme information segmented by the phrase analysis unit 420 into attributes such as subject case and target case (S105). Specifically, the sentence structure analysis unit 430, to which the sentence pattern signal is input from the phrase analysis unit 420, determines each morpheme included in the phrase based on each morpheme corresponding to the input sentence pattern signal and the phrase composed of each morpheme. Determine the “case composition”.
[0120]
That is, as shown in FIG. 5, for example, when the dependency element of each morpheme is “” or “is”, the sentence structure analysis unit 430 applies the subject (subject) to the morpheme before the dependency element. Or a leading character). For example, when the dependency element of each morpheme is “NO” or “NO”, the sentence structure analysis unit 430 determines that the morpheme before the dependency element is an object (target).
[0121]
Further, for example, when the dependency element of each morpheme is “Yes”, the sentence structure analysis unit 430 includes an action (predicate; this predicate is composed of a verb, an adjective, and the like). It is determined that
[0122]
The sentence structure analysis unit 430 that has determined the “case structure” of each morpheme that constitutes each clause determines the range of topics (topics) to be described later based on the first morpheme information associated with the determined “case structure”. A topic search command signal for specifying is output to the topic search unit 360.
[0123]
Next, the utterance type determination unit 440 performs a step of specifying an utterance type indicating the type of utterance content based on the phrase specified by the phrase analysis unit 420 (S106). Specifically, “speech sentence type” (speech type) is determined based on each morpheme corresponding to the sentence pattern signal input from the phrase analysis unit 420 and a phrase composed of each morpheme.
[0124]
That is, the utterance type determination unit 440 collates each phrase with each dictionary stored in the utterance type database 460 based on each phrase corresponding to the input sentence pattern signal, and from each phrase, Extract sentence elements related to the dictionary. The utterance type determination unit 440 that extracts sentence elements related to each dictionary from each phrase determines “spoken sentence type” based on the extracted sentence elements.
[0125]
The utterance type determination unit 440 outputs an answer search command signal for causing the corresponding user to search for a specific answer sentence to the answer sentence search unit 370 based on an instruction from the topic search unit 360 described later.
[0126]
Next, the reflexive determination unit 320 compares the first morpheme information extracted by the morpheme extraction unit 410 with each fixed content, and performs a step of searching for fixed content including the first morpheme information from each fixed content. (S107: reflective processing).
[0127]
Specifically, the reflexive determination unit 320 to which the topic search command signal is input from the sentence structure analysis unit 430 receives the first morpheme information included in the input topic search command signal and the reflection element database 801. The reflection element information (standard content) is collated, the reflection element information including the first morpheme information is searched from each reflection element information, and the searched reflection element information is output to the management unit 310.
[0128]
When the reflection element information cannot be searched for the reflection element information including the first morpheme information from the reflection element information, the reflection determination unit 320 determines whether to return the topic search command signal input from the sentence structure analysis unit 430. To the unit 330.
[0129]
Next, the return determination unit 330 collates the first morpheme information extracted by the morpheme extraction unit 410 with each return element, and performs a step of searching for the return element including the first morpheme information from each return element ( S108;
[0130]
When it is determined that the first morpheme information is included in each return element, the return determination unit 330 acquires the return element including the first morpheme information, and manages the response sentence including the acquired return element. The data is output to the unit 310 (turnback processing). That is, assuming that the return element (answer sentence output last time, utterance content uttered by the previous user, etc.) is S and the first morpheme information is W, the return determination unit 330 has a relationship of W が S and W ≠ φ. If it is established, the turn-back process shown above is performed.
[0131]
On the other hand, when determining that the first morpheme information is not included in each return element, the return determination unit 330 sends the topic search command signal input from the reflective determination unit 320 to the conversation range determination unit 340. Output.
[0132]
Then, the discourse range determination unit 340 collates the first morpheme extracted by the phrase analysis unit 420 with each discourse range, and searches (determines) the discourse range including the first morpheme information from each discourse range. A step is performed (S109).
[0133]
Specifically, the conversation range determination unit 340 to which the topic search command signal is input from the reversal determination unit 330 is based on the input search command signal, and the content that the user is speaking from the conversation database 500. Search for a relevant range (discourse range).
[0134]
For example, the discourse range determination unit 340 includes the first morpheme information included in the input topic search command signal (an interesting movie; *; is there) {is there an interesting movie? }, The first morpheme information is compared with the discourse range group, and if the discourse range group includes a morpheme (for example, “movie”) constituting the first morpheme information, The “movie” included in the information is determined as the discourse range. In this case, since the discourse range “movie” is included in the first morpheme information, the discourse range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the topic search unit 360.
[0135]
On the other hand, when the first morpheme information does not include a discourse range group, the conversation range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the abbreviated sentence complement unit 350.
[0136]
Next, the abbreviation sentence complementing unit 350 includes attributes that do not include a morpheme among the attributes (subject, object, action, etc.) constituting the first morpheme information based on the first morpheme information extracted by the phrase analysis unit 420. Perform the step of searching. Thereafter, the abbreviated sentence complementing unit 350 performs a step of adding the morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute based on the attribute not including the searched morpheme (S110; abbreviated sentence). Complement).
[0137]
Specifically, the abbreviated sentence complementing unit 350 to which the topic search command signal is input from the discourse range determination unit 340 is based on the first morpheme information included in the input discourse search command signal, and the utterance including the first morpheme information. It is determined whether or not the content is an abbreviated sentence. If the utterance content including the first morpheme information is an abbreviated sentence, the morpheme in the discourse range to which the first morpheme information belongs is added to the first morpheme information.
[0138]
For example, the abbreviation sentence complementing unit 350 has a morpheme constituting the first morpheme information included in the input topic search command signal (director; *; *) (director?) (This sentence is “what” director If it is unknown, it means an abbreviated sentence.), The conversation range previously determined by the conversation range determination unit 340 (A movie name; this A movie name is the title of the movie) The first morpheme information belonging to the first morpheme information is added to the first morpheme information ("A movie name"). Director; *; *).
[0139]
That is, assuming that the first morpheme information is W and the determined discourse range is D, the abbreviated sentence complementing unit 350 adds the discourse range D to the first morpheme information W, and uses the first morpheme information after the addition as a topic search command. It is included in the signal and output to the topic search unit 360.
[0140]
Next, the topic search unit 360 collates the first morpheme information extracted by the phrase analysis unit 420 or the first morpheme information supplemented by the abbreviated sentence complement unit 350 with each second morpheme information, and each second morpheme A step of searching for second morpheme information including morpheme constituting the first morpheme information from the information is performed (S111).
[0141]
Specifically, the topic search unit 360 to which the topic search command signal is input from the talk range determination unit 340 or the abbreviated sentence complement unit 350 is based on the first morpheme information included in the input topic search command signal. The “topic title” including the morpheme of the first morpheme information is searched from each “topic title” (second morpheme information) belonging to the discourse range determined by the determination unit 340, and the search result is used as a search result signal. The information is output to the answer sentence search unit 370 and the utterance type determination unit 440.
[0142]
For example, when the “case configuration” constituting the first morpheme information is (Sato; *; I like) {I like Sato}, the topic search unit 360, as shown in FIG. Each morpheme (Sato; *; I like) belonging to the “composition” is compared with each topic title 1-1 to 1-4 belonging to the discourse range (Sato), and from among the topic titles 1-1 to 1-4 A topic title 1-1 (Sato; *; likes) that matches (or approximates) each morpheme (Sato; *; likes) belonging to the “case composition” is searched, and this search result is used as a search result signal as a response sentence. The data is output to the search unit 370 and the utterance type determination unit 440.
[0143]
The utterance type determination unit 440 to which the search result signal is input from the topic search unit 360, based on the input search result signal, an answer search command for searching for a specific answer sentence that answers to the corresponding user. A signal (this answer search command signal also includes the determined “sentence type”) is output to the answer sentence search unit 370.
[0144]
Then, the answer sentence search unit 370 collates the utterance type of the identified user with each answer type associated with the second morpheme information based on the second morpheme information searched by the topic search unit 360, A search is made for a response type that matches the user's utterance type from each response type, and a step of acquiring an answer sentence associated with the response type based on the searched response type is performed (S112).
[0145]
Specifically, the answer sentence search unit 370, to which the search result signal from the topic search unit 360 and the answer search command signal from the utterance type determination unit 440 are input, the topic title (search result) corresponding to the input search result signal. Based on the second morpheme information) and the “speech sentence type” (speech type) corresponding to the answer search command signal, from the group of answer sentences associated with the “topic title”, An answer sentence having an answer type (this answer type means “type of answer sentence” shown in FIG. 11) matching the “sentence type” (DA, IA, CA, etc.) is searched.
[0146]
For example, when the topic title corresponding to the search result is the topic title 1-1 (Sato; *; I like) shown in FIG. 12, the answer sentence search unit 370 is associated with the topic title 1-1. Of the reply sentences 1-1 (DA, IA, CA, etc.), the reply composed of the reply type (DA) that matches the “spoken sentence type” (for example, DA; utterance type) determined by the utterance type determination unit 440. The sentence 1-1 (DA; (I also like Sato) is searched), and the searched answer sentence is output to the management unit 310 as an answer sentence signal.
[0147]
Next, the management unit 310 to which the answer sentence signal is input from the answer sentence search unit 370 outputs the input answer sentence signal to the output unit 600. In addition, the management unit 310 to which the reflection element information is input from the reflective determination unit 320 or the contents of the turn-back process from the turn-back determination unit 330, the answer sentence corresponding to the input reflection element information, the contents of the turn-back process input Is output to the output unit 600 (S113). The output unit 600 to which an answer sentence is input from the management unit 310 outputs the input answer sentence {for example, I also like Sato}.
[0148]
(Operation and effect of conversation control device and conversation control method)
According to the invention according to the present application having the above-described configuration, the reflexive determination unit 320 collates the first morpheme information constituting the utterance content uttered by the user with each fixed content stored in advance, and each fixed content from among, it is possible to find the standard content including a first morpheme information, reflexive determination unit 320, for example when the first morpheme information is a fixed content such as "Hello" includes the standard content it is possible to answer the same boilerplate content "Hello" or the like.
[0149]
In addition, when the user's utterance content is a fixed content, the reflexive determination unit 320 replies with the fixed content (greeting etc.), so that the user first contacts the conversation control device 1. You can enjoy a sense of communication.
[0150]
Further, the reversal determination unit 330 collates the current first morpheme information with the past answer sentence, and if the current first morpheme information is not included in the past answer sentence, the agreement stored in advance Since the contents can be acquired, the return determination unit 330 returns the response to the past answer sentence if the input information currently input by the user matches the past answer sentence (use) It is possible to conclude that the user has input the input information that the person is listening to the answer sentence.
[0151]
In this case, since the user has turned back the past answer sentence, the return determination unit 330 acquires the stored agreement content, and the acquired agreement content (for example, “That is right”) Can be output. As a result, if the user does not understand the meaning of the answer sentence output from the conversation control device 1, the user can listen again and listen again to the answer sentence. You can taste a feeling like this.
[0152]
Further, the reversal determination unit 330 collates the current first morpheme information with the past first morpheme information, and if the current first morpheme information is included in the past first morpheme information, the repulsion content is acquired. Therefore, when the input information input last time is included in the input information input this time, the user may repeatedly input the same content as the previous input information. Therefore, it can be determined that the user has not properly answered the answer text from the conversation control device.
[0153]
In this case, since the user does not appropriately reply to the previous response sentence, the reversal determination unit 330 acquires the repulsion content stored in order to repel the user, and the acquired repulsion content Is output. As a result, if the user does not input appropriate input information in response to the reply from the conversation control device 1, the content of repulsion is output from the conversation control device 1, so that the user can also talk with other users. You can taste a sense of being.
[0154]
Further, in order to search for the second morpheme information approximate to the first morpheme information, the topic search unit 360 may collate each second morpheme information belonging to the “discourse range” with the first morpheme information. Since it is not necessary to collate the second morpheme information with the first morpheme information, the time until the second morpheme information approximate to the first morpheme information can be shortened.
[0155]
As a result, the topic search unit 360 can search (pinpoint search) the second morpheme information that approximates the first morpheme information in a short time, so that the answer sentence search unit 370 is the topic search unit 360. An answer sentence associated with the second morpheme information can be acquired in a short time based on the searched second morpheme information, and the conversation control device 1 quickly answers the utterance content from the user. be able to.
[0156]
Moreover, the topic search part 360 searches the 2nd morpheme information which contains the morpheme (element which comprises a user's utterance content) which comprises 1st morpheme information from each 2nd morpheme information, and an answer sentence search part Since the answer sentence 370 can acquire the answer sentence associated with the second morpheme information based on the second morpheme information searched by the topic search part 360, the answer sentence search part 370 displays the utterance content of the user. Based on each constituent morpheme (first morpheme information), it is possible to take into account the semantic space (subject, subject, etc.) constructed by each morpheme and to obtain an answer sentence created in advance based on such semantic space Thus, it is possible to obtain an answer sentence more suitable for the utterance content than simply obtaining the answer sentence associated with the keyword using the entire utterance contents as a keyword.
[0157]
Further, since the topic search unit 360 searches for the second morpheme information including the first morpheme information, it is not necessary to search for the second morpheme information that completely matches the utterance content of the user, and the conversation control device 1 is developed. The developer who does this need not store in advance a huge amount of answer sentences corresponding to the utterance content that will be uttered by the user, and the capacity of the storage unit can be reduced.
[0158]
Furthermore, the answer sentence search unit 370 selects the answer type that matches the user's utterance type from the answer types (description, affirmation, location, repulsion, etc.) associated with each second morpheme information belonging to the “discourse range”. Since the answer sentence associated with the answer type can be acquired based on the searched answer type, the conversation control device 1 can determine the utterance type that constitutes the conversation content of the user, for example, the user Answers that match the user's utterance type out of multiple answer sentences based on what is simply expressed, what the user has to say, what the user describes the location element, etc. As a result, it is possible to obtain an optimal answer to the corresponding user.
[0159]
Furthermore, the answer sentence search unit 370 matches the user's utterance type from among the answer types associated with each second morpheme information belonging only to the “discourse range” searched by the discourse range determination unit 340. Since it is only necessary to search for the type (pinpoint search is possible), it is not necessary to search for the response type associated with the “all” second morpheme information and the user's utterance type one by one. It is possible to obtain an optimal answer sentence corresponding to
[0160]
Finally, the abbreviated sentence complementing unit 350 belongs to the discourse range in which the first morpheme information exists even if the first morpheme information constituting the utterance content of the user is an abbreviated sentence and is not clear as Japanese. If so, the discourse range can be added to the first morpheme information, and the first morpheme information consisting of abbreviated sentences can be complemented.
[0161]
Thus, the abbreviated sentence complementing unit 350 allows the first morpheme information so that the utterance contents constituting the first morpheme information are appropriate Japanese even if the utterance contents constituting the first morpheme information are abbreviated sentences. Since a specific morpheme (such as a morpheme constituting a discourse range) can be complemented, the topic search unit 360 is based on the first morpheme information after completion supplemented by the abbreviated sentence complement unit 350. The optimal second morpheme information related to the information can be acquired, and the answer sentence search unit 370 can select an answer sentence that is more suitable for the content of the user's utterance based on the second morpheme information acquired by the topic search unit 360. Can be output.
[0162]
As a result, even if the input information from the user is an abbreviated sentence, the conversation control device 1 does not use a function such as a neuro network or an AI intelligence, and does not mean what the abbreviated sentence means through past search results. Since the developer of the conversation control device 1 does not need to install a neuro network or AI intelligence, the system of the conversation control device 1 can be constructed more simply.
[0163]
[Example of change]
In addition, this invention is not limited to the said embodiment, The change as shown below can be added.
[0164]
(First change example)
In this modification, the conversation database 500 may be an element storage unit that stores a plurality of pieces of element information indicating the entire set group including a plurality of morpheme sets in association with the set group. Further, the morpheme extraction unit 410 collates the morpheme extracted from the character string with each set group, selects a set group including the extracted morpheme from each set group, and selects an element associated with the selected set group. Information may be extracted as first morpheme information.
[0165]
As shown in FIG. 16, some morphemes included in the character string uttered by the user are similar. For example, as shown in FIG. 16, if the element information indicating the entire group is “gift”, the “gift” is similar to a present, gift, year-end gift, mid-year gift, celebration, etc. (group). Therefore, when there is a morpheme similar to “gift” (such as the present), the morpheme extraction unit 410 can handle the similar morpheme as “gift”.
[0166]
That is, for example, when the morpheme extracted from the character string is “present”, the morpheme extraction unit 410 has “present” as the element information representing “present” as shown in FIG. “Present” can be replaced with “Gift”.
[0167]
This allows the morpheme extraction unit 410 to sort out morphemes that are similar to each other, so that the developer who develops the conversation control device can perform the first corresponding to the semantic space that is grasped from the mutually similar first morpheme information. It is not necessary to create answer sentences related to the bimorphic information and the second morpheme information one by one, and as a result, the amount of data stored in the storage unit can be reduced.
[0168]
(Second modified example)
As shown in FIG. 17, in the present modification example, the topic search unit 360 may include a ratio calculation unit 361 and a selection unit 362.
[0169]
The ratio calculation unit 361 compares the first morpheme information extracted by the morpheme extraction unit 410 with each second information, and the ratio of the first morpheme information to the second morpheme information for each second morpheme information. Is a calculation means for calculating.
[0170]
Specifically, the ratio calculation unit 361 to which the topic search command signal is input from the sentence structure analysis unit 430, based on the first morpheme information included in the input topic search command signal, as shown in FIG. One morpheme information is collated with each topic title (second morpheme information) belonging to the discourse range stored in the conversation database 500, and the first morpheme information occupies in each topic title for each topic title. Calculate the percentage.
[0171]
For example, as shown in FIG. 18, when the first morpheme information constituting the utterance sentence uttered by the user is (Sato; *; I like) {I like Sato}, the ratio calculation unit 361 Each morpheme belonging to “case composition” (Sato; *; likes) is compared with each morpheme (Sato; *; likes) included in the topic title, and each morpheme belonging to “case composition” is matched to the topic title. The ratio of (Sato; *; I like) is calculated to be 100%. The ratio calculation unit 361 performs these calculations for each topic title, and outputs each calculated ratio to the selection unit 362 as a ratio signal.
[0172]
The selection unit 362 is a selection unit that selects one second morpheme information from each second morpheme information according to the size of each rate calculated by the rate calculation unit 361 for each second morpheme information. is there.
[0173]
Specifically, the selection unit 362 to which the ratio signal is input from the ratio calculation unit 361 selects from among the ratios included in the input ratio signal (element of “case configuration” / element of “topic title” × 100). For example, a topic title with a high ratio is selected (see FIG. 18). The selection unit 362 that has selected a topic title with a high ratio outputs the selected topic title to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal. The answer sentence search unit 370 acquires an answer sentence associated with the topic title based on the topic title selected by the selection unit 362.
[0174]
Accordingly, the selection unit 362 calculates the ratio of the first morpheme information to the second morpheme information for each second morpheme information, and the size of each ratio calculated for each second morpheme information. Accordingly, one second morpheme information can be selected from each second morpheme information, so that the selection unit 362 has, for example, first morpheme information (which constitutes the user's utterance content). If second morpheme information with a large proportion of second morpheme information can be acquired from a plurality of second morpheme information groups, the second morpheme information that follows the semantic space grasped from the first morpheme information is more As a result, the answer sentence search unit 370 can make an optimum answer to the utterance content of the user.
[0175]
Further, since the selection unit 362 can select a topic title having a high ratio calculated by the ratio calculation unit 361 from a plurality of topic titles, the selection unit 362 belongs to the “case configuration” included in the user's utterance sentence. Even if each morpheme and each topic title stored in the conversation database 500 do not completely match, it is possible to acquire a topic title closely related to each morpheme belonging to “case composition”.
[0176]
As a result, since the selection unit 362 can acquire a topic title closely related to the “case configuration” constituting the first morpheme information, the developer who develops the conversation control device 1 configures the first morpheme information. Since it is not necessary to store topic titles that completely match “case configuration” in the conversation database 500, the capacity of the conversation database 500 can be reduced.
[0177]
Further, the ratio calculation unit 361 calculates the ratio of the first morpheme information to the second morpheme information for each second morpheme information belonging only to the “discourse range” searched by the discourse range determination unit 340. Therefore, it is not necessary to calculate the ratio of the first morpheme information to the “all” second morpheme information, and the second morpheme information that follows the semantic space composed of the first morpheme information is acquired in a shorter time. As a result, it is possible to quickly output an optimum answer sentence for the utterance content from the user based on the acquired second morpheme information.
[0178]
The ratio calculation unit 361 collates each morpheme of the first morpheme information belonging to each classified attribute and each morpheme of each second morpheme information belonging to each attribute stored in advance for each attribute. It may be a first search means for searching for second morpheme information including each morpheme of the first morpheme information in at least one attribute from the second morpheme information.
[0179]
Specifically, the ratio calculation unit 361 to which the topic search command signal is input, for each “case” (subject; object; action) of the “case configuration” included in the input topic search command signal. ”And each morpheme belonging to“ case ”of the topic title composed of the same“ case ”are collated, and it is determined whether or not the morphemes constituting each“ case ”are the same.
[0180]
For example, as illustrated in FIG. 19, when the “case” morpheme of “case configuration” is (dog; person; bitten) {dog bites a person}, the ratio calculation unit 361 displays those morphemes. Match "dog", "person", and "chewed" with the morpheme "dog", "person", and "chewed" of the topic title consisting of the same "case" that constitutes the morpheme Then, among the morphemes “dog”, “people”, and “chewing” that make up the topic title, the “case composition” morpheme “dog” consisting of the same “case” corresponding to each morpheme, The ratio of “people” and “biting” is calculated (100%).
[0181]
If the element constituting the topic title is (person; dog; bite) {person bites the dog}, the ratio calculation unit 361 performs morphemes belonging to two cases by the same procedure as described above. Therefore, the degree of coincidence of each “case” between the morpheme constituting “case composition” and “topic title” is calculated to be 33% (see FIG. 19).
[0182]
The ratio calculation unit 361 that has calculated the ratio selects a topic title with a high ratio from the ratios, and outputs the selected topic title to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal.
[0183]
As a result, the ratio calculation unit 361 displays, for each “case”, each morpheme of the first morpheme information belonging to each classified “case configuration” (subject case, target case, etc.) and a pre-stored topic title. Since the second morpheme information including each morpheme of the first morpheme information in at least one “case” can be searched from each topic title, the ratio calculator 361 is different from the normal word order. If the utterance content is composed of things, for example, “a person bites a dog”, the morpheme of the subject case is “person” and the morpheme of the subject case is “dog”. Search for matching second morpheme information and obtain the answer sentence {"really?" Or "I don't know the meaning"} associated with the second morpheme information (person; dog; bite) can do.
[0184]
That is, since the ratio calculation unit 361 can identify utterance contents that are difficult to identify, for example, “a person bites a dog” and “a dog bites a person”, an optimal answer to the identified utterance contents, For the former, for example, “Really?”, For the latter, for example, “Are you okay?”
[0185]
Further, the ratio calculation unit 361 may search for the second morpheme information including the morpheme of the first morpheme information in at least one attribute from the respective second morpheme information belonging to the “discourse range”. From the second morpheme information, it is no longer necessary to acquire one second morpheme information, the second morpheme information that follows the semantic space composed of the first morpheme information can be acquired in a shorter time, As a result, the conversation control device 1 can quickly output an optimal answer sentence for the utterance content from the user based on the acquired second morpheme information.
[0186]
Note that the selection unit 362 may select one topic title from among the topic titles according to a predetermined priority order. This priority order means a priority for selection as a topic title. This priority order is predetermined by the developer at the development stage.
[0187]
(Third change example)
As shown in FIG. 20, the present modification is not limited to the above-described embodiment and each modification, and data is transmitted and received via the communication network 800 in the conversation control devices 1 a and 1 b and the communication network 1000. Communication unit 900, conversation databases 500b to 500d connected to communication unit 900, and servers 2a to 2c (conversation control system).
[0188]
Here, the communication network 1000 means a communication network that transmits and receives data. In the present embodiment, for example, the Internet is used. In this modification, the conversation databases 500b to 500d and the servers 2a to 2c are limited for convenience. However, the present invention is not limited to this, and other conversation databases may be provided.
[0189]
Thereby, the conversation control unit 300 refers not only to the conversation database 500a arranged inside the conversation control device 1a but also to the other conversation control device 1b, the other conversation databases 500b to 500d, and the servers 2a to 2c. Therefore, for example, it is a case where it is not possible to search from the conversation database 500a the discourse range related to each morpheme (first morpheme information) belonging to the “case configuration” included in the topic search command signal. However, by referring to the other conversation control device 1b, conversation databases 500b to 500d, and servers 2a to 2c, it is possible to search the range of conversation related to the above “case structure” and the like based on the user's utterance sentence. A suitable answer sentence can be searched.
[0190]
(Fourth change example)
The sentence structure analysis unit 430 may store in the conversation database 500 each “case configuration” and each “case configuration” constituting the identified first morpheme information. The answer sentence search unit 370 may store each “case structure” and each “morpheme” associated with each “case structure” constituting the searched answer sentence in the conversation database 500.
[0191]
The conversation range determination unit 340 may store the searched conversation range in the conversation database 500. The topic search unit 360 may store the searched second morpheme information in the conversation database 500.
[0192]
The first morpheme information, the second morpheme information, each “case composition” that constitutes the first morpheme information or the second morpheme information, each morpheme associated with each “case composition”, and the retrieved answer sentence Each "case structure" and each morpheme associated with each "case structure" and the searched discourse range are associated with each other and stored in the conversation database 500 or the reversal element database 802 as history morpheme information. be able to.
[0193]
The abbreviated sentence complementing unit 350 does not include a morpheme from each attribute (subject, object, action, etc .; case configuration) constituting the first morpheme information based on the first morpheme information extracted by the phrase analysis unit 420. The attribute is searched, and the history morpheme information stored in the conversation database 500 or the return element database 802 is added to the attribute based on the searched attribute.
[0194]
Specifically, the abbreviated sentence complementing unit 350 to which the topic search command signal is input from the discourse range determination unit 340 is based on the first morpheme information included in the input discourse search command signal, and the utterance including the first morpheme information. If the content is an abbreviated sentence, and the utterance content made up of the first morpheme information is an abbreviated sentence (for example, the subject, object, or action does not have a predetermined morpheme), the conversation database 500 or The history form information stored in the return element database 802 is added to the first morpheme information.
[0195]
That is, assuming that the subject included in the history form information is S1, the object is O1, the action A1, the discourse range is D1, and the omitted first morpheme information is W, the supplemented first morpheme information W1 is S1ＳW , O1∪W, A1∪W, or D1∪W.
[0196]
The topic search unit 360 collates the first morpheme information W1 supplemented by the abbreviated sentence complement unit 350 and each second morpheme information, and from each “topic title” (second morpheme information), the first morpheme information The second morpheme information including W1 is searched, and the searched topic title is output to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal.
[0197]
Thereby, even if the utterance content composed of the first morpheme information is an abbreviated sentence and is not clear as Japanese, the abbreviated sentence complementing unit 350 uses the history form information stored in the conversation database 500, Since the omitted morpheme of the first morpheme information can be complemented, the utterance content including the omitted first morpheme information can be clarified.
[0198]
For this reason, when the abbreviated sentence complementing part 350 is the abbreviated sentence which comprises 1st morpheme information, 1st morpheme information is set so that the utterance content which consists of 1st morpheme information becomes appropriate Japanese. Since the morpheme omitted in FIG. 6 can be supplemented, the topic search unit 360 is based on the first morpheme information supplemented with the morpheme, and the optimal “topic title” (second morpheme) related to the first morpheme information. Information) and the answer sentence search unit 370 can output an answer sentence that is more suitable for the content of the user's utterance based on the optimal “topic title” acquired by the topic search unit 360. .
[0199]
(Fifth change example)
The topic search unit 360 may include a deletion unit 363 and a discourse addition unit 364, as shown in FIG. The deletion unit 363 collates the second morpheme information with the discourse range searched by the discourse range determination unit 340 based on the searched second morpheme information, and determines the discourse from each morpheme constituting the second morpheme information. It is a deletion means for deleting a morpheme that matches the range.
[0200]
Specifically, the topic search unit 360 to which the topic search command signal is input from the abbreviated sentence complement unit 350 receives the first morpheme information included in the input topic search command signal and the discourse determined by the discourse range determination unit 340. Each second morpheme information belonging to the range is collated, and second morpheme information matching the first morpheme information is searched from each second morpheme information.
[0201]
Then, based on the searched second morpheme information, the deletion unit 363 collates the second morpheme information with the morphemes constituting the discourse range determined by the discourse range determination unit 340, and stores the second morpheme information. Then, the morpheme that matches the morpheme constituting the discourse range is deleted, and the second morpheme information from which the morpheme has been deleted is output to the discourse adding unit 364 as a delete signal.
[0202]
That is, the deletion unit 363 removes (removes) the current discourse range D2 (this D2 is made of morpheme) determined by the discourse range determination unit 340 from each morpheme t1 constituting the second morpheme information. If the result is t2, t2 = t1-D2).
[0203]
The discourse adding unit 364 acquires another discourse range associated with the discourse range searched by the discourse range determining unit 340 based on the second morpheme information from which the morpheme has been deleted by the deletion unit 363, and the acquired other This is a discourse adding means for adding a morpheme constituting a discourse range to second morpheme information.
[0204]
Specifically, when the current discourse range D2 is related to the answer sentence K1, and the discourse range is DK, other discourses related to the answer sentence K1 or the current discourse range D2 (those that have a sibling relationship). Since the range D3 can be expressed as D3 = D2∪DK, the second morpheme information W2 after adding the morpheme constituting the other discourse range D3 can be set to W2 = t2∪D3.
[0205]
For example, each morpheme t1 constituting the second morpheme information is (A movie name; *; interesting) {A movie name is interesting? }, And the current discourse range D2 determined by the discourse range determination unit 340 is (A movie name), the deletion unit 363 first starts from each morpheme t1 (A movie name; *; interesting). The conversation range D2 (A movie name) is deleted, and the deleted result is set to t2 (*; *; interesting) (t2 = t1-D2).
[0206]
When the other conversation range D3 related to the current conversation range D2 (A movie name) is “B movie name”, the second morpheme information after adding the morpheme constituting the other conversation range D3 W2 is t2∪D3, so (B movie name; *; funny) {B movie name is interesting? }.
[0207]
Thereby, when the user's utterance content is “A movie name is interesting?”, The discourse adding unit 364 matches each morpheme (A movie name; *; interesting) constituting the user's utterance content. Second morpheme information (A movie name; *; interesting), other second morpheme information (B movie name; *; interesting) {B movie name is interesting? }, The answer sentence search unit 370 acquires the answer sentence (for example, “B movie name is interesting”) associated with the second morpheme information changed by the discourse adding part 364, The acquired answer text can be output.
[0208]
As a result, the answer sentence search unit 370 does not directly output the answer sentence to the user's utterance contents, but based on the second morpheme information including the morpheme added by the discourse addition part 364, the utterance contents Therefore, the output unit 600 can output a more humane answer sentence based on the answer sentence searched by the answer sentence search unit 370.
[0209]
Note that the discourse adding unit 364 is not limited to adding the other discourse range to the second morpheme information from which the morpheme has been deleted, but to the second morpheme information from which the morpheme has been deleted. (Stored in the database 500) may be added.
[0210]
(Sixth change example)
When the topic search unit 360 cannot search the second morpheme information including the first morpheme information from the second morpheme information, the topic search unit 360 collates the first morpheme information with each answer sentence, When the answer sentence including the first morpheme information can be searched from among the first morpheme information, the first search unit may acquire the second morpheme information associated with the searched answer sentence.
[0211]
Specifically, the topic search unit 360, to which the topic search command signal is input from the abbreviated sentence complement unit 350, based on the first morpheme information included in the input topic search command signal, If the second morpheme information that matches the first morpheme information cannot be obtained from each second morpheme information by collating with the morpheme information, as shown in FIG. The answer sentence associated with the second morpheme information is collated.
[0212]
When the topic search unit 360 determines that the morpheme constituting the first morpheme information (the action or the morpheme associated with the action) is included in the answer sentence by this collation, the answer sentence includes Search for associated second morpheme information.
[0213]
Thereby, even if the topic search unit 360 cannot search the second morpheme information that matches the first morpheme information from each second morpheme information, Since the answer sentence including the morpheme (action or the morpheme associated with the action) that constitutes the user can be identified and the second morpheme information associated with the identified answer sentence can be searched, the user's utterance content 2nd morpheme information corresponding to the 1st morpheme information which constitutes can be searched appropriately.
[0214]
As a result, since the topic search unit 360 can search for the optimal second morpheme information corresponding to the first morpheme information, the answer sentence search unit 370 can search for the optimal second morpheme information searched by the topic search unit 360. Based on the above, it is possible to obtain an appropriate answer sentence for the user's utterance content.
[0215]
[program]
The contents described in the conversation control system and the conversation control method can be realized by executing a dedicated program for using a predetermined program language in a general-purpose computer such as a personal computer.
[0216]
Here, as the programming language, the topic that the user wants, the user's emotional level for a certain matter, or the type of statement sentence, affirmative sentence, question sentence, repulsive sentence, etc. are associated with the morpheme according to the semantic content In this embodiment, for example, DKML (Discourse Knowledge Markup Language), XML (eXtensible Markup Language), C language, etc. developed by the inventors can be used.
[0217]
That is, the conversation control device 1 stores data stored in each of the conversation databases 500a to 500d (stored information such as second morpheme information, fixed content, answer sentence, answer type, set group, discourse range, element information, etc.), etc. These parts can be constructed by DKML (Discourse Knowledge Markup Language), XML (eXtensible Markup Language), or the like, and can be realized by executing a program for using the constructed storage information.
[0218]
According to such a program according to the present embodiment, each morpheme constituting the utterance content of the user is identified, the semantic content grasped from each identified morpheme is analyzed, and associated with the analyzed semantic content. Generally, a conversation control device, a conversation control system, and a conversation control method that have the effect of being able to output an optimal answer sentence corresponding to a user's utterance content by outputting a previously prepared answer sentence. It can be easily realized by a general purpose computer.
[0219]
In addition, since the developer who develops the conversation control device 1 can hierarchically construct the second morpheme information and the like for searching the answer sentence for the utterance content of the user using the language in the database. The conversation control device 1 can acquire an answer sentence to the utterance content from the database through a hierarchical procedure based on the utterance content of the user.
[0220]
In other words, the conversation control device 1 has determined and determined the hierarchy of the user's utterance content (for example, whether the second morpheme information stored in the database is in a higher concept or a lower concept). An appropriate answer sentence can be acquired from the answer sentences stored in advance based on the hierarchy.
[0221]
For this reason, the conversation control device 1 does not collate the first morpheme information composed of the user's utterance content and the previously stored “all” second morpheme information one by one. Since the second morpheme information and the first morpheme information may be collated, the second morpheme information approximate to the first morpheme information can be acquired in a short time.
[0222]
Furthermore, communication between the communication unit 800 and the communication unit 900 may be performed by transmitting and receiving data via a communication network 1000 using a protocol such as DKML. Thereby, for example, when the conversation control device 1 does not have an answer sentence suitable for the user's utterance content, the conversation control device 1 is suitable for the user's utterance content through the communication network 1000 according to the convention such as DKML. The retrieved answer sentence (described in DKML or the like) can be retrieved, and the retrieved answer sentence can be acquired (see FIG. 20).
[0223]
The program can be recorded on a recording medium. As shown in FIG. 23, examples of the recording medium include a hard disk 1100, a flexible disk 1200, a compact disk 1300, an IC chip 1400, and a cassette tape 1500. According to the recording medium on which such a program is recorded, the program can be easily stored, transported, sold, and the like.
[0224]
[Second Embodiment]
(Basic configuration of conversation control device)
A second embodiment of the present invention will be described with reference to the drawings. The conversation control device 1 in the present embodiment is substantially the same as the internal structure of the conversation control unit 300 in the first embodiment, but has a selection unit 380 (selection means) and a part of the function of the answer sentence search unit 370. It is different in different points (see FIG. 24). Since it is the same as the structure of 1st embodiment and a modified example except this difference, description about structures other than a difference is abbreviate | omitted.
[0225]
In addition, in this embodiment, it naturally has various functions in the first embodiment, for example, a turning process, a reflective process, and the like, and an effect corresponding to these functions is naturally produced. In the present embodiment, first morpheme information and topic titles described later have a “case” configuration in which each morpheme belonging to the first morpheme information and topic title is a subject, object, action, etc., as in the first embodiment. It may be included regardless.
[0226]
In the first embodiment, the process in which the conversation control device 1 outputs an answer sentence corresponding to the utterance content based on the utterance content from the user has been described. In this embodiment, a process will be described in which the conversation control device outputs an answer sentence different from the previous answer sentence when a predetermined time has elapsed since the same utterance content was first uttered. The structure of the conversation control device 1 in this embodiment is as follows. Note that the conversation control method and program in the present embodiment can be executed with the same processing content as the processing content in the conversation control device 1.
[0227]
As shown in FIG. 25, in the present embodiment, the phrase includes a topic title indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof, and each topic title includes a topic to the user. Answer sentences are associated. In this embodiment, the topic uses the topic title shown in FIG. Moreover, the priority order for selecting as an answer sentence is matched with the topic title containing a common morpheme. Each phrase, each topic title, and each answer sentence is stored in advance in the conversation database 500 (morpheme storage means).
[0228]
As shown in FIG. 25, for example, when the topic title 1 is (A device; IR (Inazuma Rush); tell me) {Tell me about the A device IR (Inazuma Rush)}, the topic title 1 is Topic titles that contain common morphemes (IR; tell me) are: Topic title 2 (Company A; IR (Investor Relations); Tell me) {Tell us about IR (Investor Relations)} 3 (Company B; IR (Investor Relations); tell me) {Tell me about Company B's IR (Investor Relations)}.
[0229]
The topic search unit 360 (first search means) searches each phrase including input information from a plurality of phrases stored in advance based on the input information input by the user. Specifically, the topic search unit 360 to which the first morpheme information is input from the discourse range determination unit 340 or the abbreviated sentence complement unit 350, the topic title including the first morpheme information based on the input first morpheme information. Search multiple times (this is included in the phrase).
[0230]
When each topic title is searched for by the topic search unit 360, the selection unit 380 (selection means) selects one topic title according to the priority level associated with each searched topic title. Is elected. Specifically, as shown in FIG. 24, when the first morpheme information (IR; tell) is extracted by the sentence structure analysis unit 430, the topic search unit 360 uses the first morpheme information (IR; teach). Topic title 1 to topic title 3 including
[0231]
Since the priority order associated with the identified topic title 1 is 1, the selection unit 380 selects the topic title 1 associated with the priority order 1 from the topic titles 1 to 3. That is, when a plurality of topic titles are searched by the topic search unit 360, the selection unit 380 specifies the highest priority among the respective priority orders associated with the searched topic titles. The topic title associated with the identified priority order is selected. Note that the selection unit 380 may specify the lowest priority among the above priorities and select a topic title associated with the specified priority.
[0232]
In addition, the selection unit 380 and the answer sentence search unit 370 are acquired first when a predetermined time has elapsed since the same input information was first acquired based on the input information acquired by the input unit 100. The answer sentence different from the answer sentence corresponding to the same input information is searched.
[0233]
That is, the selection unit 380 (selection means) is based on each phrase searched by the topic search unit 360, and the searched phrase is included in each phrase searched this time. When the time elapses, a phrase that is different from the previously searched phrase is selected from the phrases searched this time. Here, the selection unit 380 determines whether or not a predetermined time has elapsed since the previous phrase was searched, using a timer function in its own unit. The answer sentence search unit 370 (second search means) searches for an answer sentence corresponding to the phrase based on the phrase selected by the selection unit 380.
[0234]
Specifically, in the present embodiment, the selection unit 380 includes the topic title searched (or selected) last time in each topic title searched this time by the topic search unit 360, and the previous topic title is searched. When a predetermined time has elapsed since then, a topic title different from the previously searched (or selected) topic title is selected from each topic title searched this time.
[0235]
As shown in FIG. 25 and FIG. 26, for example, the selection unit 380 converts the topic title 1 (A device; IR; tell) previously searched (or selected) into the topic title 1 to topic title 3 searched this time. If the “predetermined time” has elapsed since the previous topic title 1 (A device; IR; tell me) was included, the topic title 1 (A device; Priority 1 associated with IR; tell me) is specified.
[0236]
This selection part 380 selects the topic title 2 (A company; IR; tell me) matched with the next priority 2 of the specified priority 1 from the topic titles 1-3 searched this time. . As shown in FIG. 25, the “predetermined time” corresponds to the application time associated with the priority order in this embodiment. This predetermined time can be changed by an operation instruction from the outside.
[0237]
On the other hand, if a predetermined time has not elapsed since the previous topic title 1 (A device; IR; tell me) has been searched, the selection unit 380 selects the topic title 1 (A device; IR; tell me) is used over other topic titles 2 and 3. Then, when the topic titles 1 to 3 including the topic title 1 searched last time are searched again after the predetermined time has elapsed, the selection unit 380 has priority associated with the previous topic title 1. A priority order 2 next to the order 1 is specified, and a topic title 2 associated with the specified priority order 2 is selected.
[0238]
When the topic title 2 (A company; IR; tell me) is selected by the selection unit 380, for example, the answer sentence search unit 370 selects the answer sentence 2 associated with the topic title 2 (IR of company A is IR ~) The output unit 600 outputs the answer sentence 2 (IR of company A is ~) searched by the answer sentence search unit 370.
[0239]
(Conversation control method using conversation control device)
The conversation control method by the conversation control apparatus 1 having the above configuration can be implemented by the following procedure. FIG. 27 is a flowchart showing the procedure of the conversation control method according to the present embodiment. Since the processing from (S201) to (S204) shown in FIG. 27 is the same as the processing from (S101) to (S104) in the first embodiment, detailed description thereof is omitted.
[0240]
As shown in FIG. 27, the topic search unit 360 uses the first morpheme information input from the discourse range determination unit 340 or the abbreviated sentence complement unit 350 as a topic title including the first morpheme information. A step of searching a plurality of (included) is performed (S205). When the selection unit 380 includes the topic title searched last time in each topic title searched this time by the topic search unit 360, and when a predetermined time has passed since the previous topic title was searched. Then, a topic title different from the previously searched topic title is selected from each topic title searched this time (S206, S208).
[0241]
As shown in FIG. 25 and FIG. 26, for example, the selection unit 380 converts the topic title 1 (A device; IR; tell) previously searched (or selected) into the topic title 1 to topic title 3 searched this time. If the “predetermined time” has elapsed since the previous topic title 1 (A device; IR; tell me) was included, the topic title 1 (A device; IR: tell me)) and specify priority order 1. This selection part 380 selects the topic title 2 (A company; IR; tell me) matched with the next priority 2 of the specified priority 1 from the topic titles 1-3 searched this time. .
[0242]
On the other hand, if a predetermined time has not elapsed since the previous topic title 1 (A device; IR; tell me) has been searched, the selection unit 380 selects the topic title 1 (A device; IR; tell me) is used over other topic titles 2 and 3. If the previous topic title is not included in each topic title searched this time, the selection unit 380 selects the topic associated with the highest priority from the topic titles searched this time. The title is searched (S206, S207).
[0243]
Then, when the topic titles 1 to 3 including the topic title 1 previously searched by the topic search unit 360 are searched again after the predetermined time has elapsed, the selection unit 380 sets the previous topic title 1 to The next priority order 2 after the associated priority order 1 is identified, and the topic title 2 associated with the identified priority order 2 is selected.
[0244]
Next, when the topic title 2 (A company; IR; tell me) is selected by the selection unit 380, for example, the answer sentence search unit 370 selects the answer sentence 2 (the company A's name) associated with the topic title 2 (IR is ~) is performed (S209). The output unit 600 outputs the answer sentence 2 (IR of company A is ~) searched by the answer sentence search unit 370.
[0245]
(Operation and effect of conversation control device and conversation control method)
According to the invention having the above configuration, when the topic title searched last time is included in each topic title searched this time and a predetermined time has passed since the previous topic title was searched, the selection unit 350 Identifies the priority associated with the previously searched topic title, and selects the topic title associated with the next priority of the identified priority from among the topic titles retrieved this time Can do.
[0246]
Specifically, for example, when a user utters “Tell me financial information” {First morpheme information (financial information; tell me)}, the contents of the utterance include Company A, Company B・ Financial information on multiple companies is included. Here, the fact that a plurality of topic titles including topic titles constituting the user's utterance content are searched means that the user's utterance content includes ambiguous meanings.
[0247]
In this case, the topic search unit 360 uses, for example, a topic title including first morpheme information (financial information; tell me), (A company; financial information; tell me), (B company; financial information; tell me) Will be searched. At this time, the topic search unit 360 searches the topic title (company A; financial information; tell me) last time, and when a predetermined time has passed since the previous search, the selection unit 380 is searched last time. The priority (for example, “1”) associated with the topic title (company A; financial information; tell me) is specified.
[0248]
The selection unit 380 searches for a topic title (company B; financial information; tell me) associated with the next priority (for example, “2”) of the identified priority. The answer sentence search unit 370 outputs an answer sentence (the financial information of company B is ~) associated with the searched topic title (company B; financial information; tell me).
[0249]
That is, when the user utters a plurality of the same utterance contents captured by a plurality of meanings and a predetermined time has elapsed since the utterance was first made, the answer sentence search unit 370 Output different answer sentences. As a result, if the predetermined time is set to be short, the answer sentences are changed one by one within a short time, so that the conversation control device 1 gives the user an impression that it is easy to forget the contents that have been answered before. It can be used as a device that can be made to operate.
[0250]
On the other hand, if the predetermined time is set to be long, the answer text is not changed so much, so that the conversation control device 1 can give the user the impression of persistently speaking the content that was answered before. Can be used as
[0251]
In addition, the answer sentence search unit 370 searches for the corresponding answer sentence using morphemes excluding Japanese-specific particles included in the input information (t, ni, o, etc.). Based on the semantic content that can be grasped, it is possible to more accurately search for an answer sentence created in advance. That is, the answer sentence search unit 370 can search for an answer sentence more suitable for the input information than simply searching for an answer sentence associated with the keyword using the entire input information as a keyword.
[0252]
Furthermore, since the answer sentence search unit 370 searches for topic titles including the first morpheme information, there is no need to search for topic titles that completely match the input information from the user. As a result, the developer who develops the conversation control device 1 does not need to store a huge amount of answer sentences corresponding to input information expected to be input from the user in advance, and can reduce the capacity of the storage unit. it can.
[0253]
(Example of change)
In addition, this invention is not limited to the said embodiment, The change as shown below can be added. When the same input information acquired by the input unit 100 is acquired more than the reference number, the selection unit 380 and the response text search unit 370 generate an answer text that is different from the answer text corresponding to the same input information. Search.
[0254]
In other words, the selection unit 380 includes the previously searched topic titles included in the currently searched topic titles, and if the currently searched topic titles are searched more than a predetermined reference number, the selection unit 380 A topic title different from the topic title selected is selected from the topic titles searched this time. The answer sentence search unit 370 searches for an answer sentence associated with the topic title based on the topic title selected by the selection unit 380.
[0255]
Specifically, as illustrated in FIGS. 28 and 29, the selection unit 380, for example, the topic title 1 (A device; IR; tell me) searched last time is searched from the topic title 1 to the topic title 3. The topic title 1 to topic title 3 searched this time is specified whether it has been searched in the past exceeding a predetermined reference number (for example, 5 times).
[0256]
When the number of times of search exceeds a predetermined reference number (for example, 5 times), the selection unit 380 is associated with the topic title 1 (A device; IR; tell me) searched last time. Priority 1 is specified. The selection unit 380 selects the topic title 2 (A company; IR; tell me) associated with the next priority order 2 of the identified priority order 1 from the topic title 1 to topic title 3 searched this time. To do.
[0257]
On the other hand, if the number of times of search does not exceed the reference number (for example, 5 times), the selection unit 380 sets the previously searched topic title 1 (A device; IR; tell me) as another topic title. Use with priority. Thereafter, the answer sentence search unit 370 searches for an answer sentence associated with the topic title based on the topic title selected by the selection unit 380. The answer sentence search unit 370 outputs the answer sentence searched through the management unit 310 to the output unit 600.
[0258]
In this case, even when the user utters a plurality of the same utterance contents captured in a plurality of meanings and the same utterance contents are uttered exceeding the reference number, the answer sentence search unit 370 performs the previous sentence. An answer sentence different from the answer sentence can be output. Thereby, if the reference number is set to be small, the answer sentences are changed one by one. Therefore, the conversation control device 1 can give the user an impression that it is easy to forget the contents that have been answered before. It can be used as a device.
[0259]
On the other hand, if the reference number is set to a large number, the answer text is not changed so much, so the conversation control device is a device that can give the user the impression of persistently speaking the content that was answered before. Can be used.
[0260]
It should be noted that it is desirable that the answer sentence different from the previous answer sentence includes the contents to be developed in advance. In this case, for example, when the same utterance content is uttered repeatedly, the conversation control device can output an answer sentence that guides the current topic to another topic.
[0261]
Thereby, when the same utterance content as the previous time is input and a predetermined time has passed since the input of the same utterance content, for example, a character appearing in a game (RPG, reasoning game, etc.) An answer sentence that guides the story to another story can be output.
[0262]
【The invention's effect】
As described above, according to the present invention, when a user utters a plurality of the same utterance contents captured in a plurality of meanings and a predetermined time has elapsed since the same utterance contents were first uttered. Can output an answer sentence different from the previous answer sentence.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a schematic configuration of a conversation control system according to a first embodiment.
FIG. 2 is a block diagram showing an internal structure of a conversation control unit and a sentence analysis unit in the first embodiment.
FIG. 3 is a diagram showing the contents of each morpheme extracted by a morpheme extraction unit in the first embodiment.
FIG. 4 is a diagram showing the contents of each phrase extracted by a phrase analysis unit in the first embodiment.
FIG. 5 is a diagram showing the contents of “case” specified by the sentence structure analysis unit in the first embodiment.
FIG. 6 is a diagram showing an “uttered sentence type” specified by an utterance type determining unit in the first embodiment.
FIG. 7 is a diagram showing the contents of each dictionary stored in the utterance type database in the first embodiment.
FIG. 8 is a diagram showing the contents of a hierarchical structure built inside the conversation database in the first embodiment.
FIG. 9 is a diagram showing a detailed relationship of a hierarchical structure built inside the conversation database in the first embodiment.
FIG. 10 is a diagram showing the content of a “topic title” built inside the conversation database in the first embodiment.
FIG. 11 is a diagram showing the content of “answer sentence type” associated with “topic title” built inside the conversation database in the first embodiment.
FIG. 12 is a diagram showing the contents of “topic title” and “answer text” belonging to “discourse range” built inside the conversation database in the first embodiment.
FIG. 13 is a diagram showing the contents of reflection element information stored in the reflection element database in the first embodiment.
FIG. 14 is a diagram showing the contents of a wrapping element and a morpheme of the wrapping element stored in the wrapping element database in the first embodiment.
FIG. 15 is a flowchart showing a procedure of a conversation control method according to the first embodiment.
FIG. 16 is a diagram showing utterance contents organized by a morpheme extraction unit in the first modification.
FIG. 17 is a diagram illustrating an internal configuration of a topic search unit in a second modified example.
FIG. 18 is a diagram illustrating a state in which the ratio calculation unit in the second modification collates each morpheme belonging to “case configuration” and each “topic title” for each “topic title”.
FIG. 19 is a diagram illustrating a state in which the ratio calculation unit in the second modified example collates each morpheme belonging to “each component” and each morpheme belonging to “topic title” for each “case”.
FIG. 20 is a diagram showing a schematic configuration of a conversation control system in a third modified example.
FIG. 21 is a diagram illustrating an internal configuration of a topic search unit in a fifth modification example.
FIG. 22 is a diagram illustrating a state in which the topic search unit in the sixth modification collates first morpheme information with a topic title or an answer sentence.
FIG. 23 is a diagram showing a recording medium for storing a program in the first embodiment.
FIG. 24 is a block diagram showing an internal configuration of a conversation control device according to a second embodiment.
FIG. 25 is a diagram showing the contents of “topic title” and “answer text” constructed in the conversation database in the second embodiment (part 1);
FIG. 26 is a diagram showing a state where topic titles are selected by a selection unit according to the second embodiment (part 1);
FIG. 27 is a diagram illustrating a processing procedure of a conversation control method according to the second embodiment.
FIG. 28 is a diagram showing the contents of “topic title” and “answer text” constructed in the conversation database in the second embodiment (part 2);
FIG. 29 is a diagram showing how topic titles are selected by a selection unit according to the second embodiment (part 2);
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Conversation control apparatus, 2 ... Server, 100 ... Input part, 200 ... Voice recognition part, 300 ... Conversation control part, 310 ... Management part, 320 ... Reflective determination part, 330 ... Reversal determination part, 340 ... Determination of conversation range 350, abbreviated sentence complementing part, 350 ... selection part, 360 ... topic search part, 361 ... ratio calculation part, 362 ... selection part, 363 ... deletion part, 364 ... discourse addition part, 370 ... answer sentence search part, 380 ... selection unit, 400 ... sentence analysis unit, 410 ... morpheme extraction unit, 420 ... sentence analysis unit, 430 ... sentence structure analysis unit, 440 ... speech type determination unit, 450 ... morpheme database, 460 ... speech type database, 500 ... conversation Database, 600 ... output unit, 700 ... voice recognition dictionary storage unit, 800 ... communication unit, 801 ... reflection element database, 802 ... reversal element database, 900 ... communication , 1000 ... communication network, 1100 ... hard disk 1200 ... flexible disk, 1300 ... compact disc, 1400 ... IC chip, 1500 ... Tape

Claims (4)

A conversation control device that searches for an answer sentence corresponding to input information including an utterance content input by a user,
Based on the input information input from the user, when a predetermined time has passed since the same input information was first acquired, the first input information corresponding to the same input information acquired first Having a search means for searching for an answer sentence different from the answer sentence ,
further,
The search means includes
First search means for searching for each phrase including the input information from a plurality of phrases stored in advance based on the input information input by the user;
When the phrase searched last time is included in each phrase searched this time based on each phrase searched by the first search means, and a predetermined time has elapsed since the last time the phrase was searched A selection means for selecting a phrase different from the previously searched phrase from the phrases searched this time;
A conversation control apparatus comprising: second search means for searching for the answer sentence corresponding to the phrase based on the phrase selected by the selection means. The conversation control device according to claim 1,
Morpheme extraction means for extracting at least one morpheme constituting the minimum unit of the character string as first morpheme information based on a character string included in the input information input from a user;
The phrase includes second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof, and each second morpheme information is associated with an answer sentence to the user. Morpheme storage means for storing each of the second morpheme information and each answer sentence in advance,
The first search unit includes the second morpheme information including the first morpheme information from the second morpheme information stored in advance based on the first morpheme information extracted by the morpheme extraction unit. Search multiple times,
The selection means includes, based on the second morpheme information searched by the first search means, the second morpheme information searched last time is included in the second morpheme information searched this time, When a predetermined time has elapsed since the second morpheme information was searched, the second morpheme information different from the second morpheme information searched last time is used as the second morpheme information searched this time. Selected from the inside,
The conversation control device, wherein the second search unit searches the answer sentence associated with the second morpheme information based on the second morpheme information selected by the selection unit. A conversation control method in which a computer searches for an answer sentence corresponding to input information including utterance content input by a user,
Based on the input information input by the user, when the predetermined time has elapsed from when the same input information is first acquired , the search means acquires the same input information acquired first Searching for an answer sentence different from the answer sentence corresponding to
further,
Based on the input information input by the user, when the predetermined time has elapsed from when the same input information was first acquired, the search means receives the same input acquired first The step of searching for an answer sentence different from the answer sentence corresponding to the information includes:
A step first search means, based on the input information entered by the user, looking from a plurality of phrases stored in advance, each phrase comprising an input information,
Based on each phrase searched by the first search means, the selection means includes the phrase searched last time in each of the phrases searched this time, and a predetermined time from when the previous phrase was searched If There has elapsed comprises the steps of selecting a different phrase from among the currently retrieved each phrase with the phrases previously retrieved,
Conversation control method the second search means, based on said phrase is selected in the selecting means, and having a retrieving said reply sentence corresponding to the phrase. The conversation control method according to claim 3,
The morpheme extraction means has a step of extracting at least one morpheme constituting the minimum unit of the character string as the first morpheme information based on the character string included in the input information input from the user,
The phrase includes second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof, and each second morpheme information is associated with an answer sentence to the user. The morpheme storage means stores the second morpheme information and the answer sentences in advance ,
The first search unit includes the second morpheme information including the first morpheme information from the second morpheme information stored in advance based on the first morpheme information extracted by the morpheme extraction unit. Search multiple times,
The selection means includes, based on the second morpheme information searched by the first search means, the second morpheme information searched last time is included in the second morpheme information searched this time, When a predetermined time has elapsed since the second morpheme information was searched, the second morpheme information different from the second morpheme information searched last time is used as the second morpheme information searched this time. Selected from the inside,
The second searching means, a conversation control method characterized by, based on the second morpheme information elected by the selecting means, for retrieving the answer sentence associated with said second morpheme information.

Images