JP2003345823A - Access system and access control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an access system which can analyze meaning contents from utterance contents, etc., on the basis of the utterance contents, etc., that a user uttered, obtain access information associated with the analyzed meaning contents, and access a Web page complying with a request made by the user according to the obtained access information. <P>SOLUTION: This access system is provided with: a topic retrieval part 360 which retrieves 2nd morpheme information including morphemes constituting 1st morpheme information among pieces of 2nd morpheme information by collating the 1st morpheme information constituting the utterance contents of the user with the previously stored pieces of 2nd morpheme information; and an access information retrieval part 390 which obtains access information associated with the 2nd morpheme information according to the 2nd morpheme information retrieved by the topic retrieval part 360. <P>COPYRIGHT: (C)2004,JPO

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a Web server provided from a Web server distributed on a communication network.
The present invention relates to an access system and an access control method for accessing a page.

[0002]

2. Description of the Related Art Conventionally, when a user obtains information necessary for the user, the user has to obtain the information requested by the user using a search engine built in a personal computer. it can. As a result, the user can easily and quickly acquire a literature search or the like that has been conventionally performed without going to a library or the like to search.

[0003]

However, with the development of information communication in recent years, information other than the information requested by the user can be searched by the conventional search using the Internet or the like, and the user can truly request the information. It was difficult to retrieve only information.

[0004] When a user searches for information through the Internet, the user must input information via a keyword.
URL (Uniform Resource Loc) of the site requested by the user
ator), you can refer to the content of the corresponding site, but you need to find out which site has the information you need and display the information you need immediately Could not.

[0005] On the other hand, at present, a user can search for a site associated with those keywords by inputting a plurality of keywords. According to this search, the user can perform a search (refinement search) after narrowing down the information requested by the user to a certain extent, so that the user can acquire necessary information within a certain time. appear.

However, in this refined search, it is necessary to consider each keyword related to the site requested by the user and input the keyword one by one. It took a considerable amount of time. In addition, since the URL (access information) of a site other than the site required by the user is also displayed, the user has to perform the refined search a plurality of times.

Accordingly, the present invention has been made in view of the above points, and has analyzed and analyzed semantic contents that can be grasped from the utterance contents based on the utterance contents uttered by the user. An object of the present invention is to provide an access system and an access control method that can access a site that meets a user's request based on the acquired access information by acquiring the access information related to the semantic content.

[0008]

SUMMARY OF THE INVENTION According to the invention of the present application, when accessing a Web page provided from a Web server distributed on a communication network, an input is performed based on input information input by a user. Specify a character string indicating information, based on the specified character string, extract at least one morpheme constituting the minimum unit of the character string as first morpheme information, one character, a plurality of character strings or these Second morpheme information indicating a morpheme composed of a combination;
A plurality of pieces of access information for accessing a web page provided from the web server are stored in association with each other in advance, and the extracted first morphological information is compared with each of the second morphological informations. From among them, search for second morpheme information including morphemes constituting the first morpheme information, obtain access information associated with the second morpheme information based on the searched second morpheme information, and obtain the obtained access The method is characterized in that a Web page corresponding to the access information is accessed based on the information.

According to the invention of the present application, the conversation control device is configured such that the morpheme constituting the first morpheme information from among the second morpheme information (the morpheme constituting the input information inputted by the user) Is retrieved, the access information associated with the second morpheme information is acquired based on the retrieved second morpheme information, and a Web page corresponding to the access information is acquired based on the acquired access information. Since the conversation control device can access, the conversation control device, based on each morpheme constituting the input information from the user, into the semantic space constructed by each morpheme (the semantic content grasped from the character string composed of each morpheme) Based on the access information associated with the acquired access information, the user can access the Web page associated with the acquired access information. Without input one by one considering the keywords related to the site of self-request, only a conversation like between the conversation control apparatus, it is possible to automatically obtain necessary information.

[0010] Further, the conversation control device does not simply access a Web page address associated with the keyword based on a predetermined keyword input by the user, but rather speaks the contents of the user (input information). Is associated with each morpheme based on each morpheme constituting
Since it is possible to access the address of the page, as a result, a specific W
The user can access the address of the web page, and can display a site that matches the user's request.

In the invention according to the present application, the second morphological information includes:
A plurality of pieces of access information are associated, and in each piece of access information, identification information for identifying the type of the input information, such as affirmation, negation, or repulsion, is stored in advance in association with each other, and the identified character Based on the column, a type such as affirmative, negative or repulsive constituting the input information is specified as the input type, and based on the retrieved second morphological information, the identified input type is associated with the second morphological information. Matching each specific information, searching for specific information matching the input type from each specific information, and obtaining access information associated with the specific information based on the searched specific information. Is what you do.

According to the invention of the present application, the conversation control device determines the type of the input information such as affirmative, negative or repulsive based on the character string constituting the input information input by the user. Identified as the input type, collated the identified input type with each specific information associated with the retrieved second morphological information, and searched for specific information matching the input type from among the specific information, and searched. Based on the specific information, access information associated with the specific information can be obtained.

With this, the conversation control device can provide an input type that constitutes the utterance content of the user, for example, an input type that the user simply gives an opinion, an input type that includes the impressions of the user, Based on the elements described, the access information matching the input type can be obtained from the plurality of pieces of access information, and the site that matches the user's request can be displayed.

The invention according to the present application compares the extracted first morpheme information with each second morpheme information, and for each second morpheme information, determines the ratio of the first morpheme information to the second morpheme information. Calculated, according to the magnitude of each ratio calculated for each second morpheme information, from the second morpheme information, select one second morpheme information, based on the selected second morpheme information And acquiring access information associated with the second morpheme information.

According to the invention of the present application, the conversation control device collates the extracted first morpheme information with each piece of the second morpheme information, and for each piece of the second morpheme information, For the first morphological information is calculated for the ratio, according to the magnitude of each ratio calculated for each second morphological information, from each second morphological information, one second morphological information is selected, Since the access information associated with the second morpheme information can be acquired based on the selected second morpheme information, the conversation control device can, for example, obtain the first morpheme information (elements constituting the input information of the user). ) Can obtain second morpheme information having a large proportion of the second morpheme information from each second morpheme information group, the second morpheme information following the semantic space composed of the first morpheme information is obtained. Get more accurate It can be, in the result, based on the access information associated with the second morpheme information,
It is possible to access a Web page that matches the user's request.

According to the invention of the present application, a plurality of element information items indicating the entire set group consisting of a plurality of morpheme sets are stored in advance in association with the set groups, and the morpheme extracted from the character string is collated with each set group. A set group including a morpheme is selected from each set group, and element information associated with the selected set group is extracted as first morpheme information.

According to the invention of the present application, the conversation control device collates the morpheme extracted from the character string with each set group, and selects one set group including the morpheme from each set group. Since the element information associated with the selected set group can be extracted as the first morpheme information, if the first morpheme information is, for example, “present”, “present” means, for example, a present, a gift, , Celebration, etc., the conversation control device can extract the element information associated with the set group, for example, “gift” as the first morpheme information.

Thus, the developer who develops the conversation control device can make the second morpheme information corresponding to the meaning grasped from the mutually similar first morpheme information and the answer content related to the second morpheme information similar. It is no longer necessary to create one for each first morpheme information, and as a result, the amount of data to be stored in the storage unit can be reduced, and the load on the conversation control device can be reduced. Access information required by the user can be obtained quickly.

According to the invention of the present application, each morpheme constituting the second morpheme information is classified and stored in advance for each attribute such as a subject matter composed of a nominative case and a target case composed of an object case, and the extracted first morpheme is extracted. Each morpheme of the morpheme information is classified into each attribute such as subject matter and object case, and each morpheme of the first morpheme information belonging to each classified attribute and each of the second morpheme information belonging to each attribute stored in advance A morpheme is collated for each attribute, and second morpheme information that includes the morpheme of the first morpheme information in at least one attribute is searched from each piece of the second morpheme information.

According to the invention according to the present invention, the conversation control device is configured to classify each attribute (subject, subject, etc.)
Each morpheme of the first morpheme information belonging to, and each morpheme of each second morpheme information belonging to each attribute stored in advance is compared for each attribute, from among each second morpheme information, at least one attribute Since the second morpheme information including each morpheme of the first morpheme information can be searched, the conversation control device can acquire the second morpheme information closest to the word order based on the word order constituting the input information. it can.

That is, since the conversation control device can identify input information that is difficult to identify, for example, "a person bites a dog" and "a dog bites a person", the first morpheme constituting the input information can be identified. The second morpheme information matched by the information can be acquired, and as a result, access information more suitable for the user's request can be acquired based on the acquired second morpheme information.

According to the invention of the present application, a plurality of second morpheme information is associated with input information that will be input from a user or a discourse range indicating a morpheme relevant to the answer to the user. In the second morpheme information, access information is associated with each of them, a plurality of discourse ranges are stored in advance, the extracted first morpheme information is collated with each discourse range, and the first The discourse range including the morpheme information is searched, and each second morpheme information associated with the searched discourse range is compared with the first morpheme information extracted by the morpheme extraction means. , The second morpheme information including the first morpheme information is retrieved, and the access information associated with the second morpheme information is obtained based on the retrieved second morpheme information.

According to the invention of the present application, the conversation control device searches each second morpheme information belonging to the “discussion range” with the first morpheme information in order to search for the second morpheme information similar to the first morpheme information. It is only necessary to match the morpheme information (a pinpoint search is possible), and there is no need to match the "all" second morpheme information with the first morpheme information, so that it is similar to the first morpheme information. The time until the second morpheme information is searched can be reduced.

As a result, the conversation control device can search for the second morpheme information that is similar to the first morpheme information in a short time, so that the access information associated with the searched second morpheme information can be obtained. It is possible to obtain the site in a short time and quickly display a site that meets the user's request based on the obtained access information.

In the invention according to the present application, the second morphological information includes
A plurality of access information items are associated with each other, and each access information item is associated with a priority order to be selected as access information item corresponding to the input information item input by the user. Based on the search results that are stored and retrieved, the second morpheme information including the first morpheme information is ranked from the second morpheme information according to the magnitude of the frequency of the search, and the search is performed. Based on the second morpheme information, each priority associated with the second morpheme information is compared with the rank ranked by the ranking means, and among the priorities, the priority that matches the rank is determined. And acquiring access information associated with the priority based on the acquired priority.

According to the invention of the present application, the conversation control device ranks each second morpheme information according to the frequency of searching for the second morpheme information including the first morpheme information. And based on the retrieved second morpheme information,
Each priority associated with the second morphological information is compared with the ranked rank, and from each priority,
The conversation control device can acquire the priority that matches the rank, and acquire the access information associated with the priority based on the acquired priority. If the assigned rank is the degree of emotion held by the user, access information associated with the degree of the degree of emotion can be acquired.

If a Web page of interest to the user is provided according to the priority order, the conversation control device can use the Web page if the rank ranked with reference to the search result is large. Web of particular interest to the people
The page can be displayed.

As a result, for example, if the user does not input the input information including the first morpheme information that can be searched by the conversation control device, the conversation control device will not be able to retrieve the second morpheme information that matches the first morpheme information. Will no longer be able to search for, and will set the rank to a lower rank,
You will not be able to obtain information that interests you.

For this reason, the user has to input the input information so as not to set the rank ranked by the conversation control device to be low in order to obtain information of interest to the user. While enjoying the feeling of talking with the user with care,
Information on a specific Web page can be obtained.

Further, the conversation control device can output the ranked size as the degree of emotion or the like based on the search history performed between the first morpheme information and each of the second morpheme information. Therefore, a developer who develops a conversation control device does not need to develop a complicated algorithm or a neural network or the like for causing the conversation control device to calculate the degree of emotion or the like held by the user. The development time for developing the device to be achieved can be shortened.

Incidentally, the stored information including the second morpheme information and the access information includes a topic desired by the user, a degree of emotion or a statement sentence to the user in a certain matter, an affirmative sentence, a question sentence, and a repellent sentence. And the like are preferably described in a language for hierarchically storing them in a database in association with morphemes according to their semantic contents. This language is D
It is preferably composed of KML (Discourse Knowledge Markup Language). Thereby, the developer who develops the conversation control device can construct the second morphological information and the like for searching for the answer content to the input information of the user in the database in a hierarchical manner using the language,
The conversation control device can search the database for second morpheme information and the like related to the first morpheme information based on the first morpheme information corresponding to the input information of the user through a hierarchical procedure.

That is, the conversation control device determines the hierarchy of the first morpheme information corresponding to the input information (for example, whether the second morpheme information stored in the database has a higher concept or a lower concept). , And appropriate second morpheme information can be retrieved from each piece of second morpheme information stored in advance based on the determined hierarchy.

For this reason, the conversation control device stores the first morpheme information composed of the input information of the user and the information stored in advance.
It is sufficient to match each "specific" second morpheme information belonging to a certain specific hierarchy with the first morpheme information without having to match all "second morpheme information" one by one. The second morpheme information can be searched in a short time, and as a result, a Web page corresponding to the access information can be promptly displayed based on the access information associated with the searched second morpheme information.

The second morpheme information including the first morpheme information (including the discourse range, the fixed content, the answer content, and the like.
“Abbreviated as“ second morpheme information, etc. ”) means not only the case where the second morpheme information etc. directly includes the first morpheme information, but also the case where the second morpheme information etc. includes at least one morpheme constituting the first morpheme information. It also means the case.

[0035]

DESCRIPTION OF THE PREFERRED EMBODIMENTS 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.
Is a conversation control device 1 (ranking device) according to the present embodiment.
FIG. 1 is a schematic configuration diagram of a conversation control system having a.

As shown in FIG. 1, the conversation control device 1 includes an input unit 100, a speech recognition unit 200, and a conversation control unit 300.
, A sentence analysis unit 400, a conversation database 500, an output unit 600, and a speech recognition dictionary storage unit 700.

In the present embodiment, for convenience of explanation, the description is limited to the utterance content of the user (this utterance content is a type of input information), but is not limited to the utterance content of the user. Instead, the input information may be input from a keyboard or the like. Therefore, the following "utterance contents"
The description can also be made by replacing the "utterance content" with the "input information".

Similarly, in the following description, for convenience of explanation, the description will be limited to "utterance sentence type" (utterance type). However, the present invention is not limited to this "utterance sentence type". For example, it may be “input type” indicating the type of input information input from the user. Therefore, the "utterance sentence type" (utterance type) described below can also be described by replacing "utterance type" with "input type".

The input unit 100 is an obtaining unit for obtaining input information from a user. In the present embodiment, a microphone, a keyboard, and the like are used. This input unit 100
It is also a character recognition unit that specifies a character string indicating the input information based on input information (other than voice) input from the user.

Here, the input information means characters, symbols, voices and the like input through a keyboard or the like. Specifically, the input unit 100 specifies a character string indicating the input information based on the input information (other than the voice), and uses the specified character string as a character string signal.
Output to 0. Also, the input unit 100 that acquires the utterance content from the user (this utterance content is composed of voice and is a type of input information) with a microphone or the like, the input unit 100 converts the voice constituting the acquired utterance content into an audio signal. Is output to the voice recognition unit 200.

The voice recognition unit 200 is a character recognition unit that specifies a character string corresponding to the utterance content based on the utterance content obtained by the input unit 100. Specifically, the input unit 10
The voice recognition unit 200 to which the voice signal has been input from 0 analyzes the input voice signal and specifies a character string corresponding to the analyzed voice signal using a dictionary stored in the voice recognition dictionary storage unit 700. Then, the specified character string is output to the conversation control unit 300 as a character string signal. Voice recognition dictionary storage unit 7
00 stores a dictionary corresponding to a standard voice signal.

The sentence analyzing 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. Analysis unit 420 and sentence structure analysis unit 430
Utterance type determination unit 440 and morpheme database 45
0 and an utterance type database 460.

The morpheme extraction unit 410 is a morpheme extraction means for extracting each morpheme constituting the minimum unit of a character string as first morpheme information based on the character string specified by the input unit 100 or the speech recognition unit 200. .

More specifically, the morpheme extraction unit 410 to which the character string signal has been input from the management unit 310 extracts each morpheme from the character string corresponding to the input character string signal. Here, in the present embodiment, a morpheme means a minimum unit of a word configuration represented in a character string. As a minimum unit of the word configuration, as shown in FIG. 3, for example, a part of speech such as a noun, an adjective, or a verb is given. In the present embodiment, each morpheme is represented as m1, m2,..., Ml.

That is, the morpheme extraction unit 410 converts the character string corresponding to the input character string signal into a morpheme database 4
The morphemes (m1, m2,...) Matching the morpheme group are extracted from the character string by comparing the morphemes such as nouns, adjectives, and verbs stored in advance in 50, and the extracted morphemes are extracted. Is output to the phrase analysis unit 420 as an extraction signal.

The phrase analyzer 420 includes a morpheme extractor 410
Is a conversion means for converting each morpheme into a phrase form based on each morpheme extracted in. Specifically, the morphological extraction unit 4
The phrase analyzing unit 420 to which the extracted signal is input from 10 collects the phrase into a phrase format using each morpheme corresponding to the input extracted signal.

Here, the phrase format is, in this embodiment,
In Japanese grammar, it means a sentence with an independent word or one or more attached words to the independent word, or a sentence in which a character string is divided as small as possible without breaking the meaning of the Japanese grammar. In this embodiment, the clauses are p1, p
2,... Pk.

That is, as shown in FIG. 4, the phrase analyzing unit 420 determines, based on each morpheme corresponding to the input extracted signal, a dependency element of each morpheme (for example, ga
Is extracted, and each morpheme is grouped into each phrase based on the extracted dependency elements. The phrase analysis unit 420 that compiles each morpheme into each phrase, the sentence structure analysis unit 430 and the utterance type determination unit 44 use, as a sentence pattern signal, the morphological information including each morpheme that summarizes each morpheme and each morpheme that configures each morpheme.
Output to 0.

The sentence structure analysis unit 430 includes a phrase analysis unit 420
Classification means for classifying each morpheme of the first morpheme information segmented in the above into attributes such as subject matter, object case and the like. Specifically,
The sentence structure analysis unit 430 to which the sentence pattern signal is input from the sentence pattern analysis unit 420, based on each morpheme corresponding to the input sentence pattern signal and the phrase composed of each morpheme, “case configuration” of each morpheme included in the phrase. To determine.

Here, "case composition" means a case (attribute) indicating a substantial concept in a phrase, and in the present embodiment, for example, a subject (subject case) meaning a subject / subject case , Object (target case), action (action), time (time, mood, tense, aspect, etc.)
Examples of the location include location. In the present embodiment, each morpheme associated with the three cases “case” (case configuration) of the subject, object, and action is defined as first morpheme information.

That is, as shown in FIG. 5, for example, when the dependency element of each morpheme is “” or “ha”, the sentence structure analysis unit 430 determines that the morpheme before the dependency element is Judge as the subject (subject or nominative). In addition, for example, when the dependency element of each morpheme is “” or “”, the sentence structure analysis unit 430 determines that the morpheme before the dependency element is an object (object).

Further, when the dependency element of each morpheme is “Yes”, for example, the sentence structure analysis unit 430 determines that the morpheme before the dependency element is an action (predicate; this predicate is a verb, an adjective, etc.). ).

The sentence structure analysis unit 430, which has determined the “case configuration” of each morpheme constituting each phrase, based on the first morpheme information associated with the determined “case configuration”, will be described later. Is output to the reflective judgment unit 320.

The utterance type judging section 440 includes the phrase analyzing section 42
Speech contents (input information) based on the phrase specified by 0
Is a type specifying unit that specifies an utterance type (input type) indicating the type of the utterance. Specifically, the “utterance sentence type” (utterance type) is determined based on each morpheme corresponding to the sentence pattern signal input from the phrase analysis unit 420 and the phrase composed of each morpheme.

Here, in the present embodiment, the “utterance sentence type” is a statement sentence (D; Declaratio) as shown in FIG.
n), impression sentence (I; Impression), conditional sentence (C; Conditi)
on), result sentence (E; Effect), time sentence (T; Time), location sentence (L; Location), repulsion sentence (N; Negation), positive sentence (A; Answer), question sentence (Q; Question) And so on.

The statement is a sentence composed of a user's opinion or idea. In this embodiment, the statement is shown in FIG.
As shown in the example, there is a sentence such as "I like Sato". The impression sentence means a sentence composed of impressions held by the user. The location sentence means a sentence composed of locational elements.

The result sentence means a sentence composed of a sentence including a result element for a topic. What is a time statement?
It means a sentence that includes temporal elements related to the topic.

The conditional sentence means a sentence including elements such as a premise of the topic and a condition or a reason why the topic is established when one utterance is regarded as a topic. The repelling sentence means a sentence including an element that repels a user's uttering partner. An example sentence for each "utterance sentence type" is as shown in FIG.

That is, based on each phrase corresponding to the input sentence pattern signal, the utterance type determination unit 440 compares each phrase with each dictionary stored in the utterance type database 460, and , A sentence element related to each dictionary is extracted. The utterance type determination unit 440 that extracts a sentence element related to each dictionary from each phrase determines a “speech sentence type” based on the extracted sentence element. The sentence element means the type of a sentence for specifying the type of a character string, and in the present embodiment, the sentence element includes the above-described definition phrase (which means ~).

Here, the utterance type database 460 is described.
Is a definition phrase (for example, ~) as shown in FIG.
A definition expression case dictionary with a dictionary pertaining to a, a positive case dictionary with a dictionary pertaining to positive phrases (eg, agree, agree, ping pong), a dictionary with a result phrase (eg, so, so) result representation example dictionary, greeting clause (for example, Hello) greeting dictionary with a dictionary related to, negative clause (for example, by not you say stupid, opposite) is composed of such negative cases dictionary with a dictionary related to, Each dictionary is associated with “type of utterance sentence”.

Thus, the utterance type determination unit 440 compares the phrase with each dictionary stored in the utterance type database 460, extracts a sentence element related to each dictionary from the phrase, and extracts the extracted sentence element. By referring to the type of determination associated with, the “utterance sentence type” can be determined.

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. I do.

The conversation database 500 stores a plurality of pieces of second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof, and the answer content to the user with respect to the utterance content in advance. Answer storage means (discussion storage means). The conversation database 500 is also an answer storage unit (discussion storage unit) that stores a plurality of answer types indicating the types of each answer content associated with the plurality of answer contents in advance in association with the second morpheme information.

Further, the conversation database 500 stores in advance a plurality of discourse ranges (keywords) indicating morphemes constituting ranges relevant to the input contents to be input from the user or the contents of answers to the user. It is also a discourse storage means. The discourse range (keyword) is associated with a plurality of pieces of second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof, and each second morpheme information includes an answer content to the user. Associated with each.

Further, the conversation database 500 stores answer elements (discrete storage means) that classify and store each element constituting the second morphological information into attributes such as subjects each consisting of a nominative case and target cases consisting of a purpose case. ).

As shown in FIG. 8, in the present embodiment, the conversation database 500 is, when broadly divided, the range of utterance contents likely to be uttered by the user or the range of replies to the user. It is composed of a discourse range meaning (disc course) and a topic (topic) meaning a range most closely related to the content spoken by the user. As shown in the figure, the “discussion range” is positioned as a superordinate concept of “topic” in the present embodiment.

Each discourse range can be configured to have a hierarchical structure as shown in FIG. As shown in the figure, for example, the discourse range (entertainment) of a higher concept with respect to a certain discourse range (movie) is arranged in an upper hierarchical structure, and the discourse range of a lower concept with respect to the discourse range (movie) (movie) Attributes, showing movies) can be located in a lower hierarchical structure. That is, in the present embodiment, each discourse range can be arranged at a hierarchical position where the relationship among the superordinate concept, the subordinate concept, the synonym, and the synonym with the other discourse ranges is clear.

As described above, the discourse range is composed of each topic. In the present embodiment, for example, if the discourse range is A movie name, a plurality of topics related to “A movie name” Contains.

This topic means a morpheme composed of one character, a plurality of character strings, or a combination thereof, that is, each morpheme constituting the utterance content that will be uttered by the user. In the form, subject (subject), object (target), action "case"
(Attributes). In the present embodiment, each morpheme associated with these three elements is expressed as a topic title (this topic title corresponds to a lower concept of “topic”) (second morpheme information). .

It should be noted that the topic title is not limited to each morpheme associated with the above three elements, but includes other "cases", that is, times (tense, mood, aspect, etc.) meaning time. ), A location meaning a location, a condition meaning a condition, an impression meaning an impression, an effect meaning a result, and the like.

This topic title (second morpheme information)
In the present embodiment, the information is stored in the conversation database 500 in advance, and is distinguished from the first morpheme information (derived from the utterance content uttered by the user).

For example, a topic title has a discourse range of “A
In the case of "movie name", as shown in FIG. 10, subject (A movie name), object (director), action (great) {this means "the director of A movie name is wonderful"} It is composed.

If there is no morpheme associated with “case composition” (subject, object, action, etc.) in the topic title, this part is indicated by “*” in the present embodiment.

For example, what is the $ A movie name? When the sentence of｝ is converted into a topic title (subject; object; action), what is the name of ｛A movie? In the sentence of｝, “A movie name” can be specified as the subject, but other “objects” and “actions” are not elements of the sentence, so the topic title is “subject” (A
Movie name); "object" none (*); "action" none (*) (see FIG. 10).

The answer sentence means an answer sentence (answer content) to be answered to the user. In the present embodiment, the answer sentence is associated with each topic title (second morpheme information) (FIG. 8). reference). In the present embodiment, the answer sentence is shown in FIG.
As shown in FIG. 1, in order to provide an answer corresponding to the type of utterance sentence uttered by the user, a statement (D; Declaratio
n), impression sentence (I; Impression), conditional sentence (C; Conditi)
on), result sentence (E; Effect), time sentence (T; Time), location sentence (L; Location), negative sentence (N; Negation), positive sentence (A; Answer), question sentence (Q; Question) Etc. (type of answer).

That is, each answer sentence is, as shown in FIG.
For example, it is associated with the discourse range (Sato) {lower concept; home run, higher concept; grass baseball, synonyms; panda Sato / Sato / Panda} and each topic title.

As shown in the figure, for example, the topic title 1-1 is $ (Sato; *; like): This is in the order of (subject; object; action) as described above. This order is the same in the following. If “｝”, the answer sentence 1 corresponding to the topic title 1-1 is set.
-1 is (DA; statement positive statement "I like Sato"),
(IA; affirmative sentence "I really like Sato"), (C
A: Condition affirmative sentence "Because Sato's home run is very impressive"), (EA; Result affirmative sentence "I always watch TV games where Sato appears"), (TA; Time affirmative sentence) Actually, Koshien I like it because I haven't missed five consecutive at-bats in Japan "), (LA; Positive sentence" I like the serious face when standing on the blow "), (NA; Positive repulsion sentence" People who dislike Sato I don't want to talk about, goodbye ").

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 parrot return determination unit 330, and a discourse range determination unit 3
40, an abbreviation sentence complementing unit 350, a topic searching unit 360,
An answer sentence search unit 370 is provided.

The management section 310 controls the entire conversation control section 300. Specifically, the management unit 310 to which the character string signal is input from the input unit 100 or the voice recognition unit 200 converts the input character string signal into a morpheme extraction unit 410
Output to In addition, the management unit 310 includes the answer sentence search unit 37
The answer sentence searched for “0” is output to the output unit 600.

The morphological extraction unit 41
The first morphological information extracted in step 0 is compared with each fixed content,
This is a fixed form acquisition unit that searches for fixed form contents including the first morpheme information from among the fixed form contents.

Here, the standard content means reflective element information for answering the standard content to the utterance content from the user, and this reflective element information is stored in the reflective element database 801 (the standard storage means). ) Are stored in advance.
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?" And the like "typical elements".

Specifically, the reflexive judgment section 320 to which the topic search command signal is input from the sentence structure analysis section 430 is stored in the first morphological information and the reflection element database 801 included in the input topic search command signal. The reflection element information is compared with the corresponding reflection element information, and the reflection element information including the first morphological information is searched from each reflection element information, and the searched reflection element information is output to the management unit 310.

That is, assuming that the reflection element information is D1 and the first morpheme information is W, the reflection determining unit 320 determines that W∩D1 ≠ 0
When it is determined that the relationship is established, the processing for making the above-mentioned reflexive answer is performed.

For example, when the user utters the utterance contents of “good morning”, the reflective judgment unit 320 compares the utterance contents of “good morning” with each piece of reflective element information, and from among the pieces of reflective element information. , And the reflection element information “good morning” including (coinciding with) the utterance content “good morning” is retrieved, and the retrieved reflection element information “good morning” is output to the management unit 310.

When it is not possible to search the reflection element information including the utterance content from the respective reflection element information, the reflection determining section 320 returns the topic search command signal input from the sentence structure analysis section 430 in a parallax. It outputs to the determination unit 330.

The parrot return judging section 330 includes the morpheme extracting section 4
10. The current first morpheme information extracted in step 10 is compared with the past response contents stored in the parrot return element database 802, and if the current first morpheme information is included in the past response contents, the agreement is reached. This is a fixed form acquisition unit for acquiring contents.

Here, in this embodiment, the parrot return is
The content of the user's utterance as it is (or similar content)
It means to say back. In this embodiment, the parrot return element is composed of first morpheme information or the like constituting the answer content output from the conversation control device 1 immediately before.
As shown in Fig. 4, for example, "horse is beautiful"(horse;beautiful; *), "I like Sato"(Sato;*; I like)
And the like.

The parrot return element database 802
Is also an agreement storage means for storing in advance agreement details for agreeing to input information input by the user. The agreement includes, for example, the input information previously input by the user (the last input information by the user is “A
"Are you Mr.?", The agreement includes "The director of the A movie name is Mr. S"), or "Yes,""true."

More specifically, the parrot return determining unit 330 to which the topic search command signal is input from the reflexive determination unit 320 outputs, for each parrot return element, the first morphological information and the parrot return element included in the input topic search command signal. Are compared with each other to determine whether the first morpheme information is included in the parrot return element (see FIG. 14).

When the parrot return determining unit 330 determines that the first morphological information is included in each of the parrot return elements, it obtains the contents of the agreement, and sends an answer sentence including the obtained agreement to the management unit 310. Output (parrot return processing). That is,
Assuming that the parrot return element (such as the previous answer sentence) is S and the first morpheme information is W, the parrot return determination unit 330 determines that W⊂S ≠ 0
If the relationship is established, the parrot return processing described above is performed.

For example, if the conversation control device 1
The director of A movie name is Mr. S "(Director of A movie name; Mr. S;
*) (This order is subject; object; action order; the same applies hereinafter.) After that, the user responds to the output content with "A movie name is director S?" When uttering “Director A movie name; Mr. S; *), the parrot return determining unit 330 determines the user's first morpheme information (Director A movie name; Mr. S; *) and each morpheme ( Since the director of A movie name; Mr. S; *) matches, the user concludes that he or she is parroting the answer, and obtains the memorized agreement, such as "exactly." And outputs the obtained agreement.

The parrot return judging section 330 collates the current first morphological information extracted by the morphological extracting section 410 with the past first morphological information stored in the parrot return element database 802 to determine the current first morphological information. When one morpheme information is included in the past first morpheme information, it is also a fixed form acquisition unit for acquiring the content of repulsion.

More specifically, when the user utters the utterance content “Horse is beautiful” and the conversation control device 1 outputs the content of “Horse is good with dynamism” as the answer content. If the user subsequently repeats the utterance content “horse is beautiful”, the parrot return determining unit 330 determines that each morpheme (first morphological information) {horse; beautiful; *} And the previous utterance "The horse is beautiful"
Each morpheme (first morpheme information) {horse; beautiful;
*}, The user can conclude that he / she has never heard of the answer content from the conversation control device 1 "I'm glad that the horse has a lively feeling".

In this case, since the user has not heard the answer from the conversation control device 1, the parrot return determining unit 330 acquires the stored rebound content (for example, do not repeat the same content). The acquired rebound content can be output.

On the other hand, when the parrot return determination unit 330 determines that the first morpheme information is the same as the content of the previous answer sentence or that the first morpheme information is not the same as the previous first morpheme information, the reflexive determination is made. The topic search command signal input from section 320 is output to discourse range determination section 340.

Although the parrot return determination unit 330 has described the processing in the case where the user has returned the parrot to the "contents of the conversation control device 1", the following processing can also be performed. . For example, when the output unit 600 outputs an answer sentence “Horse is beautiful”, the user responds to this answer with “Why is the horse beautiful?”, “Why is it beautiful?”, Or “Why is it?” This is the process of the parrot return determination unit 330 performed when the utterance is made.

In this case, the parrot return judging section 330 outputs the answer sentence S “Horse is beautiful” and the utterance content W from the user.
("Why is a horse beautiful? (Question)" or "Why is it beautiful? (Question)"
0, c ≠ 0 (This c indicates the utterance type of W, and this utterance type is determined by the utterance type determination unit 440 described later. The utterance type includes, for example,
Questions and the like. ), So "
Parole return processing with "conditions" (processing when the user returns a parrot with a question sentence to the answer content).

As the "conditional" parrot return process, for example, when the conversation control device 1 outputs an answer sentence "Horse is beautiful", the user asks "Why is the horse beautiful?"
When the user utters the utterance content, the parrot return determination unit 330 sends an answer sentence such as "The horse is not beautiful" to resolve the user's questions, etc.
And performs a process of outputting the obtained answer sentence to the management unit 310.

[0099] The discourse range determination unit 340 includes the phrase analysis unit 42.
This is a discourse search means for comparing the first morpheme extracted in step 0 with each discourse range and searching for a discourse range including the first morpheme information from each discourse range.

[0100] Specifically, the discourse range determination unit 340 to which the topic search command signal is input from the parrot return determination unit 330 determines the user's discourse range based on the input discourse search command signal. That is, the discourse range determination unit 340 determines whether the conversation database 500 is based on the input search command signal.
, A range (discussion range) relevant to the content spoken by the user is searched.

For example, the discourse range determination unit 340 determines that the first morphological information included in the input topic search command signal is (interesting movie; *; *) {is there any interesting movie? }, It is determined whether “movie” included in the first morpheme information is included in the discourse range group stored in the conversation database 500.

When the "movie" included in the first morpheme information is included in the discourse range group stored in the conversation database 500, the discourse range determination unit 340 determines the "movie" as the discourse range. Then, the determined discourse range is included in the topic search command signal and output to the abbreviation sentence complementing unit 350.

The discourse range determination unit 340 determines that the discourse range of “movie” included in the first morpheme information is
If it is determined that the answer is not included in the discourse range group stored in 0, the answer sentence related to the first morpheme information is obtained from the conversation database 500, and the obtained answer sentence is output to the management unit 310.

Thus, a topic search unit 360, which will be described later,
Can match each “topic title” belonging to the “discussion range” determined by the discourse range determination unit 340 with the first morpheme information specified by the sentence structure analysis unit 430, so that “all” There is no need to collate the “topic title” (second morpheme information) with the first morpheme information, and the answer sentence search unit 370 described later can reduce the time required to search for the final answer sentence.

The abbreviation sentence complementing unit 350 includes a phrase analyzing unit 420
Is an attribute search unit that searches for an attribute that does not include a morpheme from among the attributes (subject, object, action, etc.) that make up the first morpheme information based on the first morpheme information extracted in (1). In addition, the abbreviation sentence complementing unit 350
Based on the retrieved attributes, the discourse range determining unit 34 is added to the attributes.
It is also a morpheme adding means for adding morphemes constituting the discourse range searched for by 0.

Specifically, the abbreviation sentence complementing unit 350 to which the topic search command signal has been input from the discourse range determining unit 340 receives the first morpheme information based on the first morpheme information included in the input discourse search command signal. Is determined to be an abbreviated sentence, and if the utterance content consisting of the first morpheme information is an abbreviated sentence, the morpheme of the discourse range to which the first morpheme information belongs is added to the first morpheme information. .

For example, the abbreviation sentence complementing unit 350 determines that the morpheme constituting the first morpheme information included in the input topic search command signal is (supervisor; *; *) (supervisor?) Is an abbreviation because it is not known whether the director is a director), the discourse range determined by the discourse range determination unit 340 (A movie name; this A movie name is the movie title) If the first morpheme information belongs to the first morpheme information, the determined discourse range (A movie name) is added to the first morpheme information (“A
Director of movie name; *; *).

That is, assuming that the first morpheme information is W, the determined discourse range is D, and the current discourse range is d, the abbreviation sentence complementing unit 350 determines if the relation W⊂ (d∈D) holds. , The first morpheme information W is converted into W∪ (d∈D) = W ′, and the converted first morpheme information is included in the topic search command signal and output to the topic search unit 360. Abbreviated sentence complementer 350
If it determines that the utterance content corresponding to the first morpheme information included in the input topic command signal is not an abbreviation, it outputs the input topic search command signal to the topic search unit 360.

Thus, even if the first morpheme information is an abbreviated sentence and is not clear as Japanese, the abbreviation sentence complementing unit 350 can determine whether the first morpheme information belongs to a certain discourse range. The discourse range D (A movie name) is added to the first morphological information W (director; *; *), and the first morphological information is W '(director of the A movie name; *; *) {director of the A movie name What? }
Therefore, even if the utterance content of the user is an abbreviation, the abbreviation can be complemented based on the determined discourse range, and the abbreviation can be clarified. .

[0110] Therefore, the abbreviation sentence complementing unit 350 determines that the utterance content constituting the first morpheme information is in proper Japanese even if the utterance content constituting the first morpheme information is an abbreviation sentence. Since one morpheme information can be complemented with a specific morpheme, the topic search unit 360 determines the optimal “topic title” (second morpheme information) related to the first morpheme information based on the complemented first morpheme information. ) Can be obtained, and the answer sentence search unit 370 can output an answer content more suitable for the utterance content of the user based on the “topic title” acquired by the topic search unit 360.

The topic search unit 360 collates the first morpheme information extracted by the phrase analysis unit 420 or the first morpheme information complemented by the abbreviation sentence complementing unit 350 with each second morpheme information, This is a first search unit that searches the morpheme information for second morpheme information including morphemes that constitute the first morpheme information.

More specifically, the topic search section 360 to which the topic search command signal has been input from the abbreviation sentence complementing section 350 receives the discourse range determination section 340 based on the first morpheme information included in the input topic search command signal. From among the "topic titles" (second morpheme information) belonging to the discourse range determined in the above, "topic title" including the morpheme of the first morpheme information is searched, and the search result is used as a search result signal to search for an answer sentence. Part 3
70 and the utterance type determination unit 440.

For example, if the “case composition” that constitutes the first morpheme information is (Sato; *; I like) {I like Sato}, the topic search unit 360 sets Each morpheme (Sato; *; I like) belonging to the above “case composition” and each topic title 1-1 to 1-4 belonging to the discourse range (Sato)
Topic title 1-1 (Sato; *) that matches (or approximates) each morpheme (Sato; *; I like) belonging to "case composition" from among topic titles 1-1 to 1-4. ; Favorite), and outputs the search result to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal.

The utterance type determining unit 440 to which the search result signal is input from the topic search unit 360 is used to search the corresponding user for a specific answer sentence to be answered based on the input search result signal. An answer search command signal (this answer search command signal also includes the determined “utterance sentence type”) is output to answer sentence search section 370.

The answer sentence search unit 370 includes a topic search unit 360
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 in. Further, the answer sentence search unit 370 collates the utterance type of the specified 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 unit that searches for an answer type that matches the utterance type of the user from among the answer types.

More 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, sets the topic title corresponding to the input search result signal. (Based on search results;
Answer sentence group (each answer content) associated with the "topic title" based on the "second morpheme information" and the "utterance sentence type" (utterance type) corresponding to the answer search command signal
, An answer sentence having an answer type (the answer type means “answer sentence type” shown in FIG. 11) that matches the “uttered sentence type” (DA, IA, CA, etc.) is searched. .

For example, the answer sentence search unit 370 determines that the topic title corresponding to the search result is the topic title 1 shown in FIG.
-1 (Sato; *; like), the answer sentence 1-1 (DA,
IA, CA, etc.), and the “utterance sentence type” determined by the utterance type determination unit 440 (eg, DA; utterance type)
Answer sentence 1-1 (D) consisting of answer type (DA) that matches
A; (I also like Sato)), and outputs the searched answer sentence to the management unit 310 as an answer sentence signal.

The management section 310 to which the answer sentence signal has been input from the answer sentence search section 370 outputs the input answer sentence signal to the output section 600. Also, the management unit 310 to which the reflection element information from the reflexive determination unit 320 or the content of the parrot return processing from the parrot return determination unit 330 is input, the answer sentence corresponding to the input reflection element information, the content of the input parrot return processing Is output to the output unit 600.

The output unit 600 is an output unit that outputs the answer sentence obtained by the answer sentence search unit 370. In the present embodiment, for example, a speaker, a display, and the like are used. Specifically, the output unit 600 to which the answer sentence is input from the management unit 310 outputs the input answer sentence {for example, I also like Sato}.

(Conversation Control Method Using Conversation Control Apparatus) A conversation control method by the conversation control apparatus 1 having the above configuration is as follows.
It can be implemented by the following procedure. FIG. 15 is a flowchart illustrating a procedure of the conversation control method according to the present embodiment.

First, the input unit 100 performs a step of acquiring the utterance content from the user (S101). Specifically, the input unit 100 obtains a voice constituting the utterance content of the user, and uses the obtained voice as a voice signal as the voice recognition unit 200.
Output to The input unit 100 specifies a character string corresponding to the input information (other than voice) based on the input information (other than voice) input by the user, and uses the specified character string as a character string signal as a conversation controller. Output to 300.

Next, the speech recognition section 200 operates the input section 10
A step of specifying a character string corresponding to the utterance content based on the utterance content acquired in step S0 is performed (S102). More specifically, the voice recognition unit 200 to which the voice signal has been input from the input unit 100 analyzes the input voice signal, and stores a character string corresponding to the analyzed voice signal in the voice recognition dictionary storage unit 700.
, And outputs the specified character string to the conversation control unit 300 as a character string signal.

Then, the morpheme extracting section 410 performs a step of extracting each morpheme constituting the minimum unit of the character string as the first morpheme information based on the character string specified by the speech recognition section 200 (S103).

More specifically, the morpheme extraction unit 410 to which the character string signal is input from the management unit 310, outputs the character string corresponding to the input character string signal and the noun, adjective, A morpheme group such as a verb 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. .

Then, the phrase analyzing unit 420 performs a step of compiling each morpheme into a phrase format based on each morpheme extracted by the morpheme extracting unit 410 (S104). Specifically, the phrase analyzing unit 420 to which the extraction signal is input from the morphological extraction unit 410 uses the respective morphemes corresponding to the input extraction signal to combine them into a phrase format.

That is, as shown in FIG. 4, the phrase analyzing unit 420 determines, based on each morpheme corresponding to the input extracted signal, a dependency element of each morpheme (for example, ga
Is extracted, and each morpheme is grouped into each phrase based on the extracted dependency elements. The phrase analysis unit 420 that compiles each morpheme into each phrase, the sentence structure analysis unit 430 and the utterance type determination unit 44 use, as a sentence pattern signal, the morphological information including each morpheme that summarizes each morpheme and each morpheme that configures each morpheme.
Output to 0.

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 matter, object case, etc. (S
105). Specifically, the sentence structure analysis unit 430 to which the sentence pattern signal is input from the sentence analysis unit 420, based on each morpheme corresponding to the input sentence pattern signal and the phrase composed of each morpheme, determines Determine the "case composition".

That is, as shown in FIG. 5, for example, when the dependency element of each morpheme is "" or "ha", the sentence structure analysis section 430 determines that the morpheme before the dependency element is Judge as the subject (subject or nominative). In addition, for example, when the dependency element of each morpheme is “” or “”, the sentence structure analysis unit 430 determines that the morpheme before the dependency element is an object (object).

Further, if the dependency element of each morpheme is “Yes”, the sentence structure analysis unit 430 determines that the morpheme preceding the dependency element is an action (predicate; this predicate is a verb, adjective, etc.). ).

The sentence structure analysis unit 430, which has determined the “case configuration” of each morpheme constituting each phrase, based on the first morpheme information associated with the determined “case configuration”, will be described later. Is output to the topic search unit 360 to specify the range of the search.

Next, the utterance type determination unit 440 performs a step of specifying an utterance type indicating the type of the utterance content based on the phrase specified by the phrase analysis unit 420 (S10).
6). Specifically, the “utterance sentence type” (utterance type) is determined based on each morpheme corresponding to the sentence pattern signal input from the phrase analysis unit 420 and the phrase composed of each morpheme.

That is, based on each phrase corresponding to the input sentence pattern signal, the utterance type determination unit 440 checks each phrase with each dictionary stored in the utterance type database 460, and , A sentence element related to each dictionary is extracted. The utterance type determination unit 440 that extracts a sentence element related to each dictionary from each phrase determines a “speech sentence type” based on the extracted sentence element.

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. I do.

Next, the reflexive judgment section 320 collates the first morpheme information extracted by the morpheme extraction section 410 with each fixed content, and searches each fixed content for a fixed content including the first morpheme information. Steps are performed (S107; reflective processing).

More specifically, the reflexive judgment section 320 to which the topic search command signal is input from the sentence structure analysis section 430 is stored in the first morphological information and the reflection element database 801 included in the input topic search command signal. Is compared with each reflection element information (standard content), and from each reflection element information,
The reflection element information including the first morpheme information is searched, and the searched reflection element information is output to the management unit 310.

When it is not possible to search for the reflection element information including the first morpheme information from each of the reflection element information, the reflexive judgment section 320 outputs the topic search command signal input from the sentence structure analysis section 430. Is output to the parrot return determination unit 330.

Next, the parrot return judging section 330 collates the first morphological information extracted by the morphological extracting section 410 with each parrot return element, and searches for a parrot return element including the first morphological information from each parrot return element. (S1
08: parrot return processing).

When the parrot return determining section 330 determines that the first morphological information is included in each parrot return element, the parrot return element obtains a parrot return element including the first morphological information, and an answer consisting of the obtained parrot return element is obtained. The sentence is output to the management unit 310 (parrot return processing). That is, assuming that the parrot return element (response sentence output last time, the utterance content of the previous user's utterance, etc.) is S and the first morpheme information is W, the parrot return determination unit 33
0 indicates that the parrot return process described above is performed when the relationship of W⊂S ≠ 0 is established.

On the other hand, when the parrot return determining unit 330 determines that the first morphological information is not included in each parrot return element, the parrot return determining unit 330 converts the topic search command signal input from the reflective determining unit 320 into a discourse range determination. Output to the unit 340.

Then, the discourse range determination unit 340 compares the first morpheme extracted by the phrase analysis unit 420 with each discourse range, and searches the discourse range including the first morpheme information from each discourse range. The step is performed (S109).

Specifically, the discourse range deciding unit 340 to which the topic search command signal is input from the parrot return determination unit 330 is used by the conversation database 5 based on the input search command signal.
00, a range (discussion range) relevant to the content spoken by the user is searched.

For example, the discourse range determination unit 340 determines that the first morpheme information included in the input topic search command signal is (interesting movie; *; *) {Do you have an interesting movie? }, It is determined whether “movie” included in the first morpheme information is included in the discourse range group stored in the conversation database 500.

When the “movie” included in the first morpheme information is included in the discourse range group stored in the conversation database 500, the discourse range determination unit 340 determines the “movie” as the discourse range. Then, the determined discourse range is included in the topic search command signal and output to the abbreviation sentence complementing unit 350.

The discourse range determination unit 340 determines that the discourse range of “movie” included in the first morpheme information is
If it is determined that the answer is not included in the discourse range group stored in 0, the answer sentence related to the first morpheme information is obtained from the conversation database 500, and the obtained answer sentence is output to the management unit 310.

Next, the abbreviation sentence complementing unit 350 extracts a morpheme from each attribute (subject, object, action, etc.) constituting the first morphological information based on the first morphological information extracted by the phrase analyzing unit 420. Perform a step of searching for attributes not included. After that, the abbreviation sentence complementer 35
0 performs a step of adding a morpheme constituting the discourse range searched by the discourse range determination unit 340 to the attribute based on the attribute that does not include the searched morpheme (S110;
Completion of abbreviations).

[0146] Specifically, the abbreviation sentence complementing unit 350 to which the topic search command signal is input from the discourse range determining unit 340 receives the first morpheme information based on the first morpheme information included in the input discourse search command signal. Is determined to be an abbreviated sentence, and if the utterance content consisting of the first morpheme information is an abbreviated sentence, the morpheme of the discourse range to which the first morpheme information belongs is added to the first morpheme information. .

For example, the abbreviation sentence complementing unit 350 determines that the morpheme constituting the first morpheme information included in the input topic search command signal is (director; *; *) (director?) (This sentence is Is an abbreviation because it is not known whether the director is a director), the discourse range determined by the discourse range determination unit 340 (A movie name; this A movie name is the movie title) If the first morpheme information belongs to the first morpheme information, the morpheme (A movie name) of the determined discourse range is added to the first morpheme information (“A movie name”). *; *).

That is, assuming that the first morphological information is W, the determined discourse range is D, and the current discourse range is d, the abbreviation sentence complementing unit 350 determines that the relation W⊂ (d∈D) holds. , The first morpheme information W is converted into W∪ (d∈D) = W ′, and the converted first morpheme information is included in the topic search command signal and output to the topic search unit 360. Abbreviated sentence complementer 350
If it determines that the utterance content corresponding to the first morpheme information included in the input topic command signal is not an abbreviation, it outputs the input topic search command signal to the topic search unit 360.

Next, the topic search unit 360 generates the first morpheme information or the abbreviation sentence complement
The first morpheme information complemented by 50 is compared with each second morpheme information, and a step of searching the second morpheme information for second morpheme information including morphemes constituting the first morpheme information is performed. (S111).

Specifically, the topic search section 360 to which the topic search command signal has been input from the abbreviation sentence complementing section 350, based on the first morphological information included in the input topic search command signal, determines the discourse range determining section 340. From among the "topic titles" (second morpheme information) belonging to the discourse range determined in the above, "topic title" including the morpheme of the first morpheme information is searched, and the search result is used as a search result signal to search for an answer sentence. Part 3
70 and the utterance type determination unit 440.

For example, if the “case composition” that constitutes the first morpheme information is (Sato; *; I like) {I like Sato}, the topic search section 360 performs the following operations as shown in FIG. Each morpheme (Sato; *; I like) belonging to the above “case composition” and each topic title 1-1 to 1-4 belonging to the discourse range (Sato)
Topic title 1-1 (Sato; *) that matches (or approximates) each morpheme (Sato; *; I like) belonging to "case composition" from among topic titles 1-1 to 1-4. ; Favorite), and outputs the search result to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal.

The utterance type determination unit 440 to which the search result signal is input from the topic search unit 360 is used to search the corresponding user for a specific answer sentence to be answered based on the input search result signal. An answer search command signal (this answer search command signal also includes the determined “utterance sentence type”) is output to answer sentence search section 370.

Then, based on the second morpheme information searched by topic search section 360, answer sentence search section 370 determines the utterance type of the identified user and each answer type associated with the second morpheme information. A step of collating and searching for an answer type that matches the utterance type of the user from each answer type and acquiring an answer sentence associated with the searched answer type is performed (S112).

[0154] 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, sets the topic title corresponding to the input search result signal. (Based on search results;
Answer sentence group (each answer content) associated with the "topic title" based on the "second morpheme information" and the "utterance sentence type" (utterance type) corresponding to the answer search command signal
, An answer sentence having an answer type (the answer type means “answer sentence type” shown in FIG. 11) that matches the “uttered sentence type” (DA, IA, CA, etc.) is searched. .

For example, the answer sentence search unit 370 determines that the topic title corresponding to the search result is the topic title 1 shown in FIG.
-1 (Sato; *; like), the answer sentence 1-1 (DA,
IA, CA, etc.), and the “utterance sentence type” determined by the utterance type determination unit 440 (eg, DA; utterance type)
Answer sentence 1-1 (D) consisting of answer type (DA) that matches
A; (I also like Sato)), and outputs the searched answer sentence to the management unit 310 as an answer sentence signal.

Next, management section 310 to which the answer sentence signal is input from answer sentence search section 370 outputs the input answer sentence signal to output section 600. Also, the reflexive determination unit 320
The management unit 310 to which the reflection element information or the content of the parrot return processing is input from the parrot return determination unit 330 outputs an answer sentence corresponding to the input reflection element information and an answer sentence corresponding to the input content of the parrot return processing. Output to the unit 600. Output unit 600 to which an answer sentence is input from management unit 310
Is the answer sentence. For example, I also like Sato.
Is output (S113).

(Operation and Effect of Conversation Control System and Conversation Control Method) According to the present invention having the above configuration, the reflexive judging unit 320 makes the first morphological information constituting the utterance content uttered by the user. Can be compared with each of the fixed contents stored in advance, and the fixed contents including the first morphological information can be searched from among the fixed contents. Hello "in the case of fixed content, such as, the standard contents same boilerplate content and" it is possible to say hello "or the like.

When the utterance content of the user is a fixed content, the reflexive judgment unit 320 answers the fixed content (such as a greeting), so that the user first communicates with the conversation control device 1. You can enjoy the feeling of communicating with each other.

The parrot return judging section 330 compares the current first morpheme information with the past answer contents. If the current first morpheme information is not included in the past answer contents, it is stored in advance. Since the consensus content can be obtained, the parrot return determination unit 330 determines that if the input information currently input by the user matches the past answer content, the user It can be concluded that the input information is a parrot return (the user is listening again to the answer content).

[0160] In this case, the parrot return determination unit 330 obtains the stored agreement contents because the user has performed a parrot with respect to the past answer contents, and obtains the obtained agreement contents (for example, "Yes. "Etc.) can be output. Thereby, if the user does not understand the meaning of the answer content output from the conversation control device 1, the user can listen again and can make the answer content valid again.
You can feel as if you are talking with other users.

The parrot return judging section 330 compares 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, rebounds. Since the content can be acquired, the parrot return determining unit 330 repeats the same content as the previous input information when the input information input last time is included in the input information input this time. Thus, it can be determined that the user has not properly responded to the content of the response from the conversation control device.

In this case, the parrot return determination unit 330 obtains the stored repulsion contents to repel the user because the user has not properly responded to the previous response contents. Outputs the content of the repulsion. As a result, unless the user inputs appropriate input information for the answer content from the conversation control device 1, the repulsion content is output from the conversation control device 1, so that the user can talk with another user. You can enjoy the feeling of being.

In order to search for the second morpheme information that is similar to the first morpheme information, the topic search unit 360 only needs to match each of the second morpheme information belonging to the “discussion range” with the first morpheme information. , "All" second morpheme information and the first morpheme information need not be collated, so that the time required to search for second morpheme information that is similar to the first morpheme information can be reduced.

As a result, the topic search unit 360 can quickly search (pinpoint search) for the second morpheme information that is similar to the first morpheme information. The answer sentence associated with the second morpheme information can be acquired in a short time based on the second morpheme information retrieved by the unit 360, and the conversation control device 1
Can promptly respond to the content of the utterance from the user.

The topic search unit 360 searches the second morpheme information for second morpheme information including morphemes constituting the first morpheme information (elements constituting the utterance contents of the user), and Since the sentence search unit 370 can acquire the answer content associated with the second morpheme information based on the second morpheme information searched by the topic search unit 360,
The answer sentence search unit 370 considers a semantic space (subject, target, etc.) constructed by each morpheme based on each morpheme (first morpheme information) constituting the utterance content of the user, and based on the semantic space. It is possible to obtain the answer content prepared in advance, and to obtain the answer content more suitable for the utterance content than to simply use the entire utterance content as a keyword and obtain the answer content associated with the keyword. Can be.

Further, since the topic search unit 360 searches for the second morpheme information including the first morpheme information, there is no need to search for the second morpheme information that completely matches the utterance content of the user, and the conversation control unit 360 The developer who develops No. 1 does not need to previously store a huge amount of response contents corresponding to the utterance contents that will be uttered by the user, and can reduce the capacity of the storage unit.

Further, the answer sentence search unit 370 matches the utterance type of the user from the answer types (statements, affirmations, places, repulsions, etc.) associated with the respective second morpheme information belonging to the “discussion range”. Since the answer type to be searched can be searched for and the answer content associated with the answer type can be acquired based on the searched answer type, the conversation control device 1 Matched the user's utterance type from multiple answers based on what the user simply gave an opinion, what the user had, and what the user described the locational elements The content of the answer can be acquired, and a more optimal answer can be given to the corresponding user.

Furthermore, the answer sentence search unit 370 selects, from among the answer types associated with each second morpheme information belonging only to the “discussion range” retrieved by the discourse range determination unit 340,
Since it is only necessary to search for an answer type that matches the user's utterance type (pinpoint search is possible), the answer type associated with "all" second morphological information and the user's utterance type are searched one by one. This eliminates the necessity, and the optimum answer content corresponding to the type of utterance of the user can be obtained in a short time.

Lastly, the abbreviation sentence complementing unit 350 determines whether the first morpheme information constituting the utterance content of the user is an abbreviation sentence, and even if the first morpheme information is not clear in Japanese, the discourse range in which the first morpheme information is present , The discourse range can be added to the first morpheme information to complement the first morpheme information composed of abbreviations.

[0170] Thus, the abbreviation sentence complementing unit 350 operates such that even if the utterance content constituting the first morpheme information is an abbreviated sentence, the utterance content constituting the first morpheme information becomes proper Japanese. Since a specific morpheme (such as a morpheme constituting the discourse range) can be complemented to the one morpheme information, the topic search unit 360 calculates the morpheme information based on the complemented first morpheme information complemented by the abbreviation sentence complementing unit 350. The optimal second morpheme information related to the first morpheme information can be obtained, and the answer sentence search unit 370 is more suitable for the utterance content of the user based on the second morpheme information obtained by the topic search unit 360. Answer contents can be output.

As a result, even if the input information from the user is an abbreviated sentence, the conversation control device 1 uses the abbreviated sentence through past search results without using a function such as a neuro network or AI intelligence. It is possible to infer what it means, and the developer of the conversation control device 1 does not need to be equipped with a neural network and AI intelligence.
The system of the conversation control device 1 can be constructed more easily.

[Modifications] The present invention is not limited to the above embodiment, and the following modifications can be made.

(First Modification) In this modification, the conversation database 500 is an element storage means for storing a plurality of element information indicating the entire set group including a plurality of morpheme sets in association with the set group. You may. 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.

As shown in FIG. 16, some morphemes included in the character string spoken by the user are similar. For example, as shown in FIG. 16, when the element information indicating the entire set group is “gift”, the “gift” is similar to a present, a gift, a gift, a gift, a celebration, etc. (set group). Therefore, when there is a morpheme similar to “gift” (such as the above-mentioned present), the morpheme extraction unit 410 can treat the similar morpheme as “gift”.

That is, when the morpheme extracted from the character string is “present”, for example, as shown in FIG. 16, the element information representative of “present” is “gift”. Therefore, the above “present” can be replaced with “gift”.

As a result, the morpheme extraction unit 410 can sort morphemes similar to each other, so that the developer who develops the conversation control device can create a semantic space grasped from each similar first morpheme information. It is not necessary to create the corresponding second morpheme information and answer contents related to the second morpheme information one by one, and as a result, the amount of data to be stored in the storage unit can be reduced.

(Second Modified Example) As shown in FIG. 17, in this modified example, a topic calculating unit 360 may include a ratio calculating unit 361 and a selecting unit 362.

The ratio calculating section 361 is composed of the morpheme extracting section 410
The first morpheme information extracted in the above and each second information is compared,
It is a calculating means for calculating the ratio of the first morpheme information to the second morpheme information for each second morpheme information.

Specifically, as shown in FIG. 17, the ratio calculation section 361 to which the topic search command signal is input from the sentence structure analysis section 430 is based on the first morpheme information included in the input topic search command signal. Then, the first morpheme information is compared with each topic title (second morpheme information) belonging to the discourse range stored in the conversation database 500, and for each topic title, the first morpheme is included in each topic title. Calculate the ratio of information.

For example, as shown in FIG. 17, the first morpheme information constituting the utterance sentence uttered by the user is (Sato;
*; I like) {If I like Sato}, the ratio calculation section 361 calculates each morpheme (Sato; *; I like) belonging to “case composition” and each morpheme (Sato; *) included in the topic title. ;
I like it), and add “case composition” to the topic title above.
Is calculated to be 100%. The ratio calculation unit 361,
These calculations are performed for each topic title, and the calculated ratios are output to the selection unit 362 as ratio signals.

The selecting unit 362 determines the ratio of each ratio calculated for each second morpheme information by the ratio calculating unit 361.
It is a selecting means for selecting one piece of second morphological information from each piece of second morphological information.

More specifically, the selection unit 362 to which the ratio signal is input from the ratio calculation unit 361 outputs the ratio (the element of “case composition” / the element of “topic title” × 100) included in the input ratio signal. For example, a topic title having a high ratio is selected from among (see FIG. 18). The selecting unit 362 that has selected a topic title having a high ratio outputs the selected topic title to the answer sentence searching unit 370 and the utterance type determining unit 440 as a search result signal. The answer sentence search unit 370 includes the selection unit 3
Based on the topic title selected in 62, an answer sentence associated with the topic title is obtained.

Thus, the selecting unit 362 calculates the ratio of the first morpheme information to the second morpheme information for each second morpheme information, and calculates the ratio of each ratio calculated for each second morpheme information. Depending on the size, from among the second morpheme information,
Since one piece of second morpheme information can be selected, the selection unit 362 may, for example, select the second morpheme information in which the first morpheme information (constituting the utterance content of the user) accounts for a large proportion of the second morpheme information. Can be obtained from the plurality of second morpheme information groups, the second morpheme information following the semantic space grasped from the first morpheme information can be obtained more accurately, and as a result, Sentence search unit 370
Can give an optimal answer to the utterance content of the user.

The selecting unit 362 can select a topic title having a high ratio calculated by the ratio calculating unit 361 from a plurality of topic titles. Can be obtained even if the morphemes belonging to "" and the topic titles stored in the conversation database 500 do not completely match.

As a result, since the selecting section 362 can acquire a topic title closely related to the “case composition” constituting the first morpheme information, the developer developing the conversation control device 1 can Since there is no need to store topic titles that completely match the “case configuration” to be formed in the conversation database 500, the capacity of the conversation database 500 can be reduced.

Further, for each piece of second morpheme information belonging only to the “discussion range” retrieved by the discourse range determination section 340, the ratio calculation section 361 calculates the ratio of the first morpheme information to the second morpheme information. Since the calculation is performed, it is not necessary to calculate the ratio of the first morpheme information to “all” of the second morpheme information, and the second morpheme information following the semantic space composed of the first morpheme information can be calculated in a shorter time. As a result, it is possible to quickly output the optimum answer content to the utterance content from the user based on the acquired second morphological information.

The ratio calculating section 361 compares each morpheme of the first morpheme information belonging to each classified attribute with each morpheme of each second morpheme information belonging to each attribute stored in advance for each attribute. Then, the first morpheme information may be a first retrieval unit that retrieves second morpheme information that includes each morpheme of the first morpheme information in at least one attribute from each second morpheme information.

More specifically, the ratio calculation unit 361 to which the topic search command signal has been input is used for each “case” (subject; object; action) of the “case configuration” included in the input topic search command signal. Each morpheme belonging to the "case" is compared with each morpheme belonging to the "case" of the topic title having the same "case" to determine whether or not the morphemes constituting the "case" are the same.

For example, as shown in FIG. 19, when the morpheme of the “case” of the “case configuration” is (dog; person; bite) {the dog bites the person}, as shown in FIG. The morphemes "dog", "people", and "bite" of the topic title consisting of those morphemes "dog", "person", and "bite", and "case" that is the same as "case" that constitutes these morphemes Of the morphemes “dog”, “person”, and “bite” that make up the topic title, the “morpheme of the“ case composition ”consisting of the same“ case ”as the“ case ”corresponding to each morpheme” Calculate (100%) the percentage that matches “dog”, “person”, and “bite”.

If the element constituting the topic title is (person; dog; bite) {people bit the dog},
Since the morphemes belonging to the two cases are different according to the same procedure as described above, the ratio calculation unit 361 determines the degree of coincidence between the morpheme constituting the “case configuration” and the “topic title” for each “case” by 33.
% (See FIG. 19).

The ratio calculating section 361 which has calculated the ratio selects a topic title having a high ratio from among the ratios, and uses the selected topic title as a search result signal as an answer sentence searching portion 37.
0 and the utterance type determination unit 440.

As a result, the ratio calculation unit 361 determines each morpheme of the first morpheme information belonging to each of the classified “case constituents” (subject case, target case, etc.) and a topic title stored in advance by each “case case”. , And the second morpheme information including each morpheme of the first morpheme information in at least one “case” from among the topic titles. If the utterance content is different from that of the utterance, for example, "a person bites a dog", the subject morpheme is "people" and the target morpheme is "dog". The second morpheme information that matches "case" can be searched, and the answer contents associated with the second morpheme information (person; dog; chew) {"really?" Or "I do not understand the meaning" } Can be obtained.

That is, the ratio calculation unit 361 determines the utterance contents that are difficult to identify, for example, “a person bites a dog” and “a dog bites a person”.
Can be identified, so that the answer best suited to the identified utterance content can be, for example, "really?" For the former and "OK" for the latter.

Further, the ratio calculating section 361 sets the "discussion range"
From the second morpheme information belonging to at least one attribute, the second morpheme information including the morpheme of the first morpheme information in at least one attribute may be searched. There is no need to acquire dimorphological information, and the second morphological information following the semantic space composed of the first morphological information can be acquired in a shorter time. As a result, the conversation control device 1 The optimum answer to the utterance content from the user can be quickly output based on the second morphological information.

(Third Modification) As shown in FIG. 20, the present modification is not limited to the above-described embodiment and each of the above-described modifications, and is not limited to the communication unit 800 in the conversation control devices 1a and 1b. Communication unit 9 for transmitting and receiving data to and from communication unit 800 via communication network 1000
00, each of the conversation databases 500b to 500d connected to the communication unit 900, and the servers 2a to 2c (conversation control system).

Here, the communication network 1000 is
This means a communication network for transmitting and receiving data. In the present embodiment, for example, the Internet is used.

In this modification, the conversation controllers 1a and 1b, 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. May be provided. The conversation databases 50 are stored in the servers 2a to 2c.
Contents similar to the contents stored in 0a to 500d are stored.

As a result, the conversation control unit 300 operates the conversation database 50 located inside the conversation control device 1a.
0a, and via the communication network 1000,
Other conversation control devices 1b, conversation databases 500b to 5b
00d, the servers 2a to 2c can also be referred to, for example, from the conversation database 500a, the discourse range associated with each morpheme (first morpheme information) belonging to the "case configuration" included in the topic search command signal Even if it is not possible to search for, the other conversation control device 1b, conversation databases 500b to 500d, servers 2a to 2c
By referring to, the discourse range related to the first morpheme information can be searched, and an answer sentence more suitable for the user's uttered sentence can be searched.

(Fourth Modification) The sentence structure analysis unit 430 stores, in the conversation database 500, each “case configuration” constituting the specified first morphological information and each morpheme associated with each “case configuration”. It may be something. The answer sentence search unit 370 may store, in the conversation database 500, each "case configuration" constituting the searched answer sentence and each morpheme associated with each "case configuration".

The discourse range determination section 340 may store the searched discourse range in the conversation database 500. The topic search unit 360 may store the searched second morphological information in the conversation database 500.

The first morpheme information, the second morpheme information, each “case configuration” constituting the first morpheme information or the second morpheme information, and each morpheme associated with each “case configuration” were searched. The respective morphemes associated with each "case composition" and each "case composition" constituting the answer sentence and the searched discourse range can be stored in the conversation database 500 as history morpheme information by associating them with each other. it can.

[0202] The abbreviation sentence complementing section 350 includes a phrase analyzing section 420.
Based on the first morphological information extracted in the above, search for an attribute that does not include a morpheme from among the attributes (subject, object, action, etc .; case composition) constituting the first morphological information, and based on the searched attribute The history morpheme information stored in the conversation database 500 may be added to the attribute.

More specifically, the abbreviation sentence complementing unit 350 to which the topic search command signal has been input from the discourse range determining unit 340 receives the first morpheme information based on the first morpheme information included in the input discourse search command signal. It is determined whether the utterance content consisting of is an abbreviation sentence, and if the utterance content consisting of the first morpheme information is an abbreviation sentence (for example, the subject, object, or action does not have a predetermined morpheme),
The history form information stored in the conversation database 500 is added to the first morpheme information.

That is, if 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 complemented first morpheme information W1 is S1∪W, O1∪
W, A1 @ W, or D1 @ W.

The topic search unit 360 includes an abbreviation sentence complementing unit 350
The first morpheme information W1 complemented by the above is collated with each second morpheme information, and the second morpheme information including the first morpheme information W1 is searched from each “topic title” (second morpheme information). 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.

Thus, even if the utterance content consisting of the first morpheme information is an abbreviated sentence and is not clear as Japanese, the abbreviation sentence complementing unit 350 allows the conversation database 50
Since the morpheme of the omitted first morpheme information can be complemented using the history morphological information stored in 0,
The utterance content including the omitted first morpheme information can be clarified.

For this reason, when the utterance content forming the first morpheme information is an abbreviation,
Since the morpheme omitted in the first morpheme information can be complemented so that the utterance content composed of the first morpheme information becomes proper Japanese, the topic search unit 360 generates the first morpheme information in which the morpheme is complemented. Based on the best "topic title" associated with the first morpheme information (second morpheme information)
Can be obtained, and the answer sentence search unit 370 can output an answer content more suitable for the utterance content of the user based on the optimal “topic title” acquired by the topic search unit 360.

(Fifth Modification) The topic search unit 360 is configured as shown in FIG.
As shown in FIG. 1, a deletion unit 363 and a discourse addition unit 364 may be provided. The deletion unit 363 performs the second morpheme information and the discourse range determination unit 34 based on the retrieved second morpheme information.
This is a deletion unit that collates the discourse range retrieved with 0 and deletes a morpheme that matches the discourse range from among the morphemes constituting the second morpheme information.

[0209] Specifically, the topic search section 360 to which the topic search command signal is input from the abbreviation sentence complementing section 350, determines the first morpheme information included in the input topic search command signal and the discourse range determination section 340. The second morpheme information is compared with the second morpheme information belonging to the discourse range, and the second morpheme information that matches the first morpheme information is searched from the second morpheme information.

[0210] Then, based on the retrieved 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. From the information, 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 deletion signal.

That is, the deletion unit 363 removes the current discourse range D2 (this D2 is composed of morphemes) determined by the discourse range determination unit 340 from each morpheme t1 constituting the second morpheme information. (Assuming that the removed result is t2, t2 = t1-D2).

The discourse adding unit 364 obtains another discourse range associated with the discourse range searched by the discourse range determination unit 340 based on the second morpheme information from which the morpheme has been deleted by the deletion unit 363, and obtains it. This is a discourse adding means for adding morphemes constituting another discourse range to the second morpheme information.

More specifically, assuming that the discourse range in which the current discourse range D2 is related to the answer sentence K1 is DK, the answer sentence K
1 or another discourse range D3 having a relationship (brother relationship) with the current discourse range D2 can be expressed as D3 = D2∪DK, so that a morpheme constituting another discourse range D3 is added. The second morpheme information W2 after the
= T2∪D3.

For example, each morpheme t1 constituting the second morpheme information is (A movie name; *; interesting) {A movie name is interesting? }, And if the current discourse range D2 determined by the discourse range determination unit 340 is (A movie name), the deletion unit 3
63, first, each morpheme t1 (A movie name; *; interesting)
From the discourse range D2 (A movie name), and the deleted result is set as t2 (*; *; interesting) (t2 = t1-D
2).

If the other discourse range D3 related to the current discourse range D2 (A movie name) is “B movie name”, the second discourse range after adding another morpheme constituting the other discourse range D3 is added. Since the dimorphological information W2 is t2∪D3, (B movie name; *; interesting) {Is the B movie name interesting? }.

Thus, if the utterance content of the user is "A movie name is interesting?"
Each morpheme (A movie name; *;
Interesting) and the second morphological information (A movie name; *; interesting), and other second morphological information (B movie name; *; interesting) {Is the B movie name 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 unit 364, The obtained answer sentence can be output.

As a result, the answer sentence search unit 370 does not output an answer sentence to the utterance content of the user,
Since the answer sentence related to the utterance content can be output based on the second morpheme information including the morpheme added by the discourse adding unit 364, the answer content to the utterance content of the user is stored in the conversation database 500. Even if there is no answer sentence, it is possible to acquire the answer content related to the utterance content of the user, and the output unit 600 outputs a more human-like answer sentence based on the answer sentence searched by the answer sentence search unit 370. Can be output.

Note that the discourse adding unit 364 is not limited to only adding the other discourse range to the second morpheme information from which the morpheme has been deleted, and may add the history morpheme to the second morpheme information from which the morpheme has been deleted. Information (stored in the conversation database 500) may be added.

(Sixth Modification) When the topic search unit 360 cannot search the second morpheme information including the first morpheme information from each of the second morpheme information, the topic search unit 360 The first search for obtaining the second morpheme information associated with the searched answer content when the content and the answer content including the first morpheme information can be searched from each answer content It may be a means.

Specifically, the topic search section 360 to which the topic search command signal has been input from the abbreviation sentence complementing section 350 outputs the first morpheme information and the first morpheme information based on the first morpheme information included in the input topic search command signal. When the second morpheme information is compared with the second morpheme information, and the second morpheme information that matches the first morpheme information cannot be obtained from the second morpheme information, FIG.
As shown in FIG. 2, the first morpheme information is compared with the answer sentence associated with the second morpheme information.

By this collation, if the topic search unit 360 determines that the morpheme (action or morpheme associated with the action) constituting the first morpheme information is included in the answer sentence, The second morpheme information associated with the answer sentence is searched.

Thus, even if the topic search unit 360 cannot search the second morpheme information from each second morpheme information for the second morpheme information that matches the first morpheme information, Since it is possible to specify an answer sentence including a morpheme (action or a morpheme associated with an action) constituting one morpheme information and search for second morpheme information associated with the specified answer sentence, the user Can be appropriately searched for the second morpheme information corresponding to the first morpheme information constituting the utterance content.

As a result, the topic search unit 360 can search for the optimal second morpheme information corresponding to the first morpheme information.
Based on the optimal second morpheme information retrieved in step (1), it is possible to acquire an appropriate answer to the utterance content of the user.

[Program] The contents described in the above conversation control system and conversation control method can be realized by executing a dedicated program for utilizing a predetermined program language on a general-purpose computer such as a personal computer.

In this embodiment, as the programming language, in the present embodiment, the topic desired by the user, the degree of emotion of the user in a certain matter, or a statement, a positive sentence, a question sentence,
A language for associating types of repulsive sentences and the like with morphemes according to their semantic contents and hierarchically accumulating them in a database, for example, DKML (Discourse Knowle) developed by the inventors.
dge Markup Language), XML (eXtensible Markup L)
anguage), C language, and the like.

That is, the conversation control device 1 stores data (second morpheme information, fixed contents, answer sentences, answer types, sets, discourse ranges, element information, etc.) stored in each of the conversation databases 500a to 500d. ), Other parts,
It can be realized by constructing with DKML (Discourse Knowledge Markup Language) or the like, and executing a program for using the constructed storage information or the like.

According to such a program according to the present embodiment, each morpheme constituting the utterance content of the user is specified, the semantic content grasped from each specified morpheme is analyzed, and the analyzed semantic content is obtained. A conversation control device, a conversation control system, and a conversation control method which have an effect of outputting an associated pre-created response content to output an optimal response content corresponding to the utterance content of the user. Can be easily realized by a general-purpose computer.

Further, the developer who develops the conversation control device 1 can hierarchically construct the second morphological information and the like for retrieving the answer content to the utterance content of the user in the database using the language. Since it is possible, the conversation control device 1 can acquire the answer content to the utterance content from the database through a hierarchical procedure based on the utterance content of the user.

That is, the conversation control device 1 determines the hierarchy of the utterance contents of the user (for example, whether the second morpheme information stored in the database is in the upper concept or the lower concept). In addition, an appropriate answer content can be obtained from each answer content stored in advance based on the determined hierarchy.

For this reason, the conversation control device 1 does not collate the first morpheme information composed of the utterance content of the user with the “all” second morpheme information stored in advance, and does Since the second morpheme information belonging to the first morpheme information may be collated with the first morpheme information, the second morpheme information approximate to the first morpheme information can be obtained in a short time.

Further, communication between the communication section 800 and the communication section 900 is performed via
Data may be transmitted and received by a protocol such as KML. Thereby, the conversation control device 1 can transmit the DK through the communication network 1000, for example, when the conversation control device 1 has no answer content suitable for the utterance content of the user.
According to the rules such as ML, it is possible to search for an answer content (written in DKML or the like) suitable for the utterance content of the user and acquire the searched answer content (see FIG. 20).

The program can be recorded on a recording medium. As shown in FIG. 23, this recording medium includes, for example, a hard disk 1100, a floppy (registered trademark) disk 1200, a compact disk 1300,
IC chip 1400, cassette tape 1500, and the like. According to the recording medium on which such a program is recorded, storage, transportation, sale, and the like of the program can be easily performed.

[Second Embodiment] (Basic Configuration of Information Processing System) A second embodiment of the present invention will be described with reference to the drawings. FIG. 24 shows the internal structure of the information processing system (conversation control system) according to the present embodiment. As shown in the figure, the information processing system is a conversation control device 1 according to the first embodiment.
It is almost the same as the internal structure of the (information processing device), except that the conversation control unit 300 has a ranking unit 380 (ranking means). Except for this difference, the structure is the same as that of the first embodiment and the modified example, and the description of the structure other than the difference will be omitted.

In the first embodiment, the process in which the conversation control device 1 acquires the optimum answer sentence corresponding to the input information based on the input information from the user has been described. The ranking section 380 at 300
A process of ranking the degree of emotion or the like held by a user based on input information from the user will be described. The specific description is as follows.

[0235] The ranking section 380 ranks the degree of emotion or the like held by the user based on the input information input by the user.
As shown in (1), it has an emotion state determination unit 381, an understanding state determination unit 382, and an opposition state determination unit 383. Here, the emotion level means the level of the emotion that the conversation control device 1 has for the input information. This emotion level is
In the present embodiment, the emotional state is determined uniquely by the emotional state determination unit 381 described later based on the input information.

[0236] The ranking section 380 determines the degree of emotion or the like held by the user based on the "discussion event" table stored in the discourse database. here,
As shown in FIG. 25, in the “Discourse Event” table, in the present embodiment, a cohesion factor (coher) indicating that the first morpheme information is a factor having relevance (cohesion) with the discourse range.
e) Topic change (shift) indicating that the first morpheme information is a factor not having relevance to the current discourse range.

Specifically, as shown in FIG. 26, in the present embodiment, the binding factor (-) and the binding factor (+) are used.
Consists of As shown in the figure, the unity factor (-) is searched by the topic search unit 360 for a topic title in which the first morphological information belongs to the previous discourse range and is associated with "low emotion level" belonging to the topic title. Means that Here, “low emotion level” means that the conversation control apparatus 1 sets a low emotion level for the topic title (for example, when the conversation control apparatus 1 is not very interested in the topic title). Means that

[0238] As shown in the figure, the unity factor (+) is that the topic title associated with "high emotion level" in which the first morphological information belongs to the previous discourse range, and belongs to the discourse range. Means that was searched. here,
“High emotion level” means that the conversation control apparatus 1 sets a high emotion level for the topic title (for example, when the conversation control apparatus 1 is interested in the topic title). I do.

As shown in FIG. 26, these unity factors (−) and (+) are based on the result of the “conditional” parrot return process by the parrot return determination unit 330 and the result of the abbreviation sentence completion process by the abbreviation sentence complementing unit 350. , Ranking unit 380.

The topic change is performed as shown in FIG.
Topic change (-) and topic change (+). The topic change (-) means that the current topic title searched by the topic search unit 360 has been changed to another topic title associated with "low emotion level" later, as shown in FIG. .

The topic change (+) indicates that the current topic title retrieved by the topic retrieval section 360 is changed to another topic title associated with “high emotion level” later, as shown in FIG. Means Change these topics (-)
(+) Indicates the discourse range determining unit 340 as shown in FIG.
Ranking section 380 based on the search result of the discourse range by
Is determined by

The “Discourse Range Event” table is
In addition to the above, an interrupt factor indicating that the first morpheme information is a factor interrupting the current topic (interruput), an esoteric factor indicating that the first morpheme information is more difficult information than the second morpheme information ( harass), an opposition factor (oppos) indicating that the first morpheme information is a contradictory factor to the second morpheme information
e) etc.

As shown in FIG. 24, the cause of the interruption is the parrot returning process by the
It is determined by the ranking unit 380 based on the reflective processing by 0. The esoteric factor is based on the search result of the topic title (second morpheme information) by the topic search unit 360,
It is determined by the ranking unit 380. The conflict factor is determined by the ranking unit 380 based on the determination of the utterance type by the utterance type determination unit 440.

[0244] The emotion state determination unit 381
A ranking unit that ranks the second morpheme information including the first morpheme information from each second morpheme information based on the search result based on 0 according to the frequency of the search.

The emotional state determination unit 381 is also a ranking means that ranks a low rank when the frequency of search is low, and ranks a high rank when the frequency of search is high. The magnitude of this rank can be expressed as the degree of emotion for the user, the degree of understanding of the input information, and the degree of conflict felt by the user.

In the present embodiment, the degree of emotion determined by the emotion state determination unit 381 can be ranked in six stages as shown in FIG. The degree of emotion is positioned in a direction in which the numerical value shown in the figure increases as the degree of emotion with respect to the user increases (in a direction of improvement).

[0247] Therefore, a transition of the degree of emotion from e (0) to e (3) means that the degree of emotion held by the user increases. On the other hand, a transition of the degree of emotion from e (0) to e (-2) means that the degree of emotion held by the user is reduced. Of the six stages,
In the present embodiment, the reference emotion level is e (0).

In the present embodiment, ranking of the degree of emotion is performed using negative emotion factors, positive emotion factors, and topic change (+) (-). In the present embodiment, the negative emotion factors include the interruption factors, the esoteric factors, the conflict factors, the unity factors (-), and the topic changes (-) described above. The emotion plus factor is composed of a unity factor (+) and a topic change (+).

In the figure, the emotion level is changed by two or more steps due to the topic change (-) (+). This is because the emotion level is "very low" or "emotional" from the current topic title. When the topic title associated with the degree “very high” is selected, the degree of emotion is changed by two or more levels.

Specifically, when the topic search unit 360 does not retrieve the second morpheme information including the first morpheme information, the emotion state determination unit 381 ranks the degree of emotion held by the user into a lower rank. I do.

That is, the fact that the second morpheme information including the first morpheme information is not retrieved by the topic retrieval section 360 means that the first morpheme information constituting the input information inputted by the user is stored in advance in each of the first morpheme information. More difficult than dimorphological information (ha
rass), the emotion state determination unit 381 determines that the input information (first morpheme information) input by the user is an emotion minus factor (difficulty factor), Set the degree of emotion to the user to be lower.

On the other hand, when the topic search unit 360 searches for the second morpheme information including the first morpheme information, the emotion state determination unit 381 ranks the degree of emotion held by the user at a higher rank. .

That is, the fact that the topic search unit 360 has searched for the second morpheme information including the first morpheme information means that the first morpheme information constituting the input information input by the user is stored in advance. Since it is possible to artificially determine that the input information is easier than the second morpheme information, the emotional state determination unit 381 determines that the input information input by the user can be understood, and Set the degree of emotion to be held higher.

The emotional state judging section 381 is also a ranking means for ranking based on the judgment result by the discourse range determining section 340 (relevance judging means). In the present embodiment, the discourse range determination unit 340 compares the first discourse range indicating the searched discourse range with the second discourse range indicating the discourse range searched thereafter, and determines that the second discourse range is the first discourse range. This is a relevance determining means for determining whether or not the range is relevant.

More specifically, when the discourse range determination unit 340 determines that the second discourse range is not related to the first discourse range, the emotional state determination unit 381 holds the user, for example. Rank the degree of emotion lower.

That is, the fact that the discourse range determination unit 340 determines that the second topic range is not related to the first topic range means that the current discourse range including the first morpheme information (second discourse range) Can be artificially determined to be different from the previous discourse range (first discourse range) (there is no united surname in the contextual topic; this is also called "topic change"). If the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”,
First morpheme information is negative emotion factor (change topic (-))
Is determined, and the degree of emotion held by the user is set to the lower one (see FIGS. 26 and 27).

As a result, the topic change (-) means that the current discourse range is changed to another discourse range, and the first morpheme information belonging to the changed discourse range is not interesting to the conversation control device 1. It means there is.

In this case, if the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “high emotion level”, It is also possible to determine that the morpheme information is an emotion plus factor (topic change (+)) and set the degree of emotion to the user to be higher.

As a result, in the topic change (+), the current discourse range is changed to another discourse range, and the first morpheme information belonging to the changed discourse range is an interesting content for the conversation control device 1. Means that.

On the other hand, when the discourse range determination unit 340 determines that the second discourse range is related to the first discourse range, the emotion state determination unit 381 determines, for example, the degree of emotion held by the user. Rank higher.

That is, the fact that the discourse range determination unit 340 has determined that the second topic range has relevance to the first topic range means that the current discourse range including the first morpheme information (second discourse range) Since it can be quasi-determined that they are substantially the same as the previous discourse range (first discourse range) (there is a cohesive topical context), the emotional state determination unit 381 determines If the first morpheme information belonging to the two discourse range is included in the topic title (second morpheme information) associated with “high emotion level”, the first morpheme information is expressed as an emotion plus factor (cohesion factor (+)). It is determined that there is, and the degree of emotion held by the user is set higher (see FIGS. 26 and 27).

As a result, the unity factor (+) indicates that the current discourse range is substantially the same as the previous discourse range, and the first morpheme information belonging to the current discourse range is an interesting content for the conversation control device 1. Means that

In this case, if the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”, It is also possible to determine that the morpheme information is a negative emotion factor (cohesion factor (-)) and set the degree of emotion to the user to be lower.

As a result, the unity factor (-) is such that the current discourse range is substantially the same as the previous discourse range, but the first morpheme information belonging to the current discourse range is the conversation control device 1.
Means that the content is not interesting.

The emotional state judging section 381 is also a ranking means for ranking the morphemes constituting the discourse range in the abbreviation sentence complementing section 350 based on the result of the addition to the first morphological information.

More specifically, the abbreviation sentence complementing section 350 selects one of the attributes (the nominative case composed of the nominative case, the target case composed of the nominative case, etc.) constituting the first morpheme information extracted by the sentence structure analyzing section 430. , Search for attributes that do not include morphemes. The abbreviation sentence complementing unit 350 adds, based on the retrieved attribute, a morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute.

[0267] The emotion state determination unit 381 is the abbreviation sentence complementing unit 3.
When the morphemes constituting the discourse range are added to the first morpheme information at 50, the degree of emotion held by the user is ranked higher.

That is, in the abbreviation sentence complementing unit 350, the input information consisting of the first morpheme information is replaced with the abbreviation sentence (when the discourse range is added to the first morpheme information, the first morpheme information before the morpheme is added) Means that the input information consisting of is abbreviated sentence), the user assumes that the input information belonging to the current discourse range belongs to the current discourse range. Since the input information is considered to have been input, the input information input at this stage is likely to be an abbreviated sentence omitting the morpheme consisting of the current discourse range (details are described in the first embodiment. (Refer to the section of “abbreviated sentence complementing section 350”).

When the abbreviation sentence complementing unit 350 determines that the first morpheme information is an abbreviation, the emotion state determination unit 381 sets the current discourse range to be substantially the same as the previous discourse range. The first morpheme information is determined to be a cohesion factor having relevance to the previous discourse range. In this case, if the unity factor is the unity factor (+), the emotion state determination unit 381 sets the degree of emotion to the user to be higher, and if the unity factor is the unity factor (-), It is possible to set a lower degree of emotion for the user.

Further, based on the search result by the reflexive judgment unit 320 or the parrot return judgment unit 330, the emotional state judgment unit 381 determines the fixed content including the first morpheme information or the frequency of the search for the parrot return element. It is also a ranking means. The reflexive determination unit 320 is a fixed form acquisition unit that compares the extracted first morpheme information with each fixed content and searches each fixed content for reflective element information including the first morpheme information. Specifically, when the reflexive determination unit 320 searches for the fixed content that matches the first morpheme information from among the fixed content, the emotion state determination unit 381
Ranks the degree of emotion held by the user at a lower rank.

That is, the fact that the reflexive judgment section 320 has searched for the fixed content that matches the first morpheme information is an element (interrupt factor) for interrupting the current topic by the first morpheme information. Therefore, the emotional state determination unit 381 determines that the first morpheme information is a negative emotional factor (interruption factor), and sets the emotional level to the user to be lower. (Fig. 26,
See FIG. 27). For example, if the user inputs “Good morning” (regular content) that is completely unrelated to the topic during the conversation, “Good morning” becomes a cause of interruption.

The parrot return judging section 330 determines whether the current first morphological information extracted by the morphological extracting section 410 is the past answer content extracted by the morphological extracting section 410 (this past first morphological information (parrot returning element ) Is stored in the parrot return element database 802).

When the parrot return determining unit 330 determines that the current first morpheme information is included in the past answer contents, the emotional state determining unit 381 returns the user to the past That the first morpheme information is an interruption factor for interrupting the current topic.

In this case, since the current first morpheme information is an emotion minus factor (interruption factor), emotion state determination section 381 sets the degree of emotion to the user to be lower. (See FIGS. 26 and 27).

Further, the parrot return judging section 330 determines that the current morphological information extracted by the morphological extracting section 410 is past first morphological information (this past first morphological information is stored in the parrot returning element database 802). If the emotion state determination unit 381 determines that the current first morpheme information is included in the past first morpheme information,
Regardless of the current topic, it can be determined that the user has repeatedly input the same input information as the past input information.

[0276] In this case, emotion state determination section 381 determines that the input information repeatedly input is the cause of the interruption, and sets the degree of emotion held by the user to a lower level.
See “Parrot return processing” in the first embodiment).

If the parrot return judging section 330 or the reflexive judging section 320 cannot search the fixed form contents that match the first morpheme information from each of the fixed form contents, the emotion state determining section 3
81 may rank the degree of emotion held by the user at a higher rank.

Further, when the input information specified by the utterance type determination unit 440 is a repulsion element such as a repulsion sentence, the emotion state determination unit 381 ranks the user with a lower degree of emotion in the lower rank. Also attached.

[0279] Here, the repulsion element means an element for the user to repel the answer content output from the conversation control device 1, and in the present embodiment, for example,
It consists of a repulsive sentence indicating that the input information of the user is a repellent sentence to the answer content, a negative sentence indicating that the input information of the user is a negative sentence to the answer content, and the like. is there.

That is, the fact that the input type of the input information input by the user is specified by the utterance type determination unit 440 to be of a repulsive element means that the user and the conversation control device 1 are in opposition (oppose). ), The emotional state determination unit 381 determines that the first morphological information is an emotion minus factor (conflict factor), and determines the degree of emotion held by the user. It is set to the lower one (see FIGS. 26 and 27).

The comprehension state judgment section 382 includes the topic search section 36
If the second morpheme information including the first morpheme information is not retrieved from each of the second morpheme information at 0, it is determined that the input information input by the user is difficult (harass) (difficulty factor), This is a ranking means for ranking the understanding of the input information to a lower rank.

That is, the fact that the second morpheme information including the first morpheme information is not searched for from the second morpheme information by the topic search unit 360 means that the first morpheme information is more difficult than the second morpheme information. Since it can be simulated, the understanding state determination unit 382 determines that the first morpheme information is a cause of difficulty and sets the understanding level of the input information to a lower level.

When the topic search unit 360 searches for the second morpheme information including the first morpheme information from the second morpheme information, the understanding state determination unit 382 determines whether the input has been input by the user. It may be determined that the information is easy, and the understanding of the input information may be ranked higher. When the first morpheme information is a factor that causes a topic change to a cohesion factor, the understanding state determination unit 382 may rank the understanding level of the input information to a higher rank.

Here, the degree of understanding means the degree of understanding of the input information inputted by the user. As shown in FIG. 28, this understanding can be ranked in four stages in the present embodiment.

The larger the numerical value (the direction from r (−2) to r (1)), the more the input information is understood (the direction toward clearer understanding). . On the other hand, the smaller the numerical value is (r (1) to r
(-2) means that the input information is in a direction in which it is less understood (in a direction in which the understanding is unclear).

[0286] Specifically, as shown in the figure, the understanding state determination unit 382 determines that the first morpheme information is the cause of difficulty and determines the degree of understanding to be lower, based on r (0). When the first morpheme information is determined to be a factor that changes the topic to a cohesion factor, the level of understanding is changed to a higher level.

When the understanding state determination unit 382 determines that the first morpheme information is a cause of interruption or a confrontation, the first morpheme information does not contribute to the understanding of the input information. Is maintained as it is (see the figure). Furthermore, when the emotion level shifts in the positive direction, the understanding state determination unit 382 shifts the rank of the understanding level to a higher rank.

[0288] The difficulty factors, unity factors, topic changes, interruption factors, and confrontation factors are the same as those described in the section of the emotional state determination unit 381, and thus description thereof will be omitted.

If the input information specified by the utterance type judgment unit 440 is a repulsion element (oppose; also called an opposition factor) such as a repulsion sentence, the confrontation state judgment unit 383 judges the degree of opposition with the user. Is ranked higher. In addition, the confrontation state determination unit 383 includes the utterance type determination unit 4
If the input information specified in 40 is not a repelling element such as a repelling sentence, the degree of conflict with the user is ranked lower.

Here, the degree of conflict means the degree of conflict that the conversation control device 1 has for the user.
As shown in FIG. 29, in this embodiment,
It can be ranked in five stages. The higher the numerical value, the lower (weaker) the degree of conflict with the user. On the other hand, it means that the smaller the numerical value, the higher (stronger) the degree of confrontation with the user.

[0291] Specifically, as shown in the figure, when the first morphological information is determined to be a conflicting factor, the conflicting state determining unit 383 determines that the degree of conflict is lower, based on O (0). And transition (with rank). Also, the conflict state determination unit 383
Determines that the first morphological information is not a contradictory factor but a factor that changes the discourse into a cohesive factor (not a repulsive factor), and shifts the confrontation to a higher level.

Further, when the emotion level for the first morpheme information is set in a higher direction by the emotion state determination section 381, the conflict state determination section 383 sets the conflict level to a lower level (weaker the conflict level). can do. When the emotion level for the first morpheme information is set in a lower direction by the emotion state determination unit 381, the confrontation state determination unit 383
The degree of conflict can be set higher (the degree of conflict is stronger).

When the first morpheme information is an interruption factor or an esoteric factor, the conflict state determination unit 383 determines the rank of the conflict degree because the first morpheme information does not repel the answer content. Maintain the status quo (see figure).

The emotion state determination section 381, the understanding state determination section 382, or the conflict state determination section 383 outputs the determined emotion level, understanding level, and conflict level to the output section 600, and the output section outputs the input emotion level. Display degree, understanding, and degree of conflict.

(Information Processing Method Using Information Processing System) An information processing method by the information processing system having the above configuration can be implemented by the following procedure. FIG.
0 is a flowchart showing the procedure of the ranking method according to the present embodiment.

As shown in the figure, first, the input unit 100
Performs a step of acquiring input information input by the user (S201). Then, the morpheme extraction unit 410
A step of extracting each morpheme constituting the input information based on the input information acquired by the input unit 100 is performed (S2).
02).

After that, the sentence structure analysis unit 430 performs a step of compiling each morpheme into a phrase form based on each morpheme extracted by the morpheme extraction unit 410 (S203). In the present embodiment, each morpheme belonging to one phrase format is the first morpheme information.

Next, the reflexive judgment section 320 performs a step of performing reflexive processing based on the first morpheme information specified by the sentence structure analysis section 430 (S204, S20).
5). Then, the parrot return determining unit 330 performs a step of performing a parrot return process or a conditional parrot return process based on the first morpheme information specified by the sentence structure analysis unit 430 (S206).

Thereafter, the discourse range determining unit 340 performs a step of collating the first morpheme information with each discourse range and searching for a discourse range including the first morpheme information from each discourse range (S207). Further, if the input information consisting of the first morpheme information is an abbreviation, the abbreviation sentence complementing unit 350 performs a step of adding a morpheme consisting of a discourse range to the first morpheme information (S208).

[0300] Then, the topic search unit 360 collates the first morpheme information with each piece of the second morpheme information based on the specified first morpheme information or the first morpheme information to which the morpheme is added. A step of searching the morpheme information for second morpheme information including the first morpheme information is performed (S20).
9). Since the specific processing of S201 to S208 is the same as the content of the “conversation control method” described in the first embodiment, the details of these processing are omitted.

Next, the emotional state judging section 381 determines whether to use the reflexive judging section 320, the parrot returning judging section 330, the discourse range determining section 340, the abbreviation sentence complementing section 350 or the topic searching section 360 based on the search result. The step of ranking the degree of emotion to be held is performed (S210). Specifically, when the topic search unit 360 does not search for the second morpheme information including the first morpheme information, the emotion state determination unit 381 ranks the degree of emotion held by the user at a lower rank.

That is, the fact that the topic search unit 360 does not search for the second morpheme information including the first morpheme information means that the first morpheme information constituting the input information inputted by the user is stored in advance. More difficult than dimorphological information (ha
rass), the emotion state determination unit 381 determines that the input information (first morpheme information) input by the user is an emotion minus factor (difficulty factor), Set the degree of emotion to the user to be lower.

On the other hand, when the topic search unit 360 searches for the second morpheme information including the first morpheme information, the emotion state determination unit 381 ranks the degree of emotion held by the user at a higher rank. .

That is, the fact that the second morpheme information including the first morpheme information has been searched by the topic search section 360 means that the first morpheme information constituting the input information input by the user is stored in advance. Since it is possible to artificially determine that the input information is easier than the second morpheme information, the emotional state determination unit 381 determines that the input information input by the user can be understood, and Set the degree of emotion to be held higher.

[0305] Further, the emotional state determination unit 381 can also perform a step of ranking the degree of emotion held by the user based on the determination result by the discourse range determination unit 340. In the present embodiment, the discourse range determination unit 340 compares the first discourse range indicating the searched discourse range with the second discourse range indicating the discourse range searched thereafter, and determines that the second discourse range is the first discourse range. It is determined whether or not there is a relationship with the range.

[0306] Specifically, when the discourse range determination unit 340 determines that the second discourse range is not related to the first discourse range, the emotion state determination unit 381 determines the emotion that the user has. Rank degrees to lower ranks.

That is, the fact that the discourse range determination unit 340 determines that the second topic range is not related to the first topic range means that the current discourse range including the first morpheme information (second discourse range) Can be artificially determined to be different from the previous discourse range (first discourse range) (there is no united surname in the contextual topic; this is also called "topic change"). If the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”,
First morpheme information is negative emotion factor (change topic (-))
Is determined, and the degree of emotion held by the user is set to the lower one (see FIGS. 26 and 27).

As a result, the topic change (-) means that the current discourse range is changed to another discourse range, and the first morpheme information belonging to the changed discourse range is not interesting to the conversation control device 1. It means there is.

In this case, if the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “high emotion level”, It is also possible to determine that the morpheme information is an emotion plus factor (topic change (+)) and set the degree of emotion to the user to be higher.

As a result, the topic change (+) changes the current discourse range to another discourse range, and the first morpheme information belonging to the changed discourse range is an interesting content for the conversation control device 1. Means that.

[0311] On the other hand, when the discourse range determination unit 340 determines that the second discourse range is related to the first discourse range, the emotion state determination unit 381 increases the degree of emotion held by the user. Rank in the rank.

That is, the fact that the discourse range determination unit 340 has determined that the second topic range has relevance to the first topic range means that the current discourse range including the first morpheme information (second discourse range). Since it can be quasi-determined that they are substantially the same as the previous discourse range (first discourse range) (there is a cohesive topical context), the emotional state determination unit 381 determines If the first morpheme information belonging to the two discourse range is included in the topic title (second morpheme information) associated with “high emotion level”, the first morpheme information is expressed as an emotion plus factor (cohesion factor (+)). It is determined that there is, and the degree of emotion held by the user is set higher (see FIGS. 26 and 27).

As a result, the unity factor (+) is such that the current discourse range is substantially the same as the previous discourse range, and the first morpheme information belonging to the current discourse range is interesting for the conversation control device 1. Means that

In this case, if the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”, It is also possible to determine that the morpheme information is a negative emotion factor (cohesion factor (-)) and set the degree of emotion to the user to be lower.

As a result, the unity factor (-) is such that although the current discourse range is substantially the same as the previous discourse range, the first morpheme information belonging to the current discourse range is the conversation control device 1.
Means that the content is not interesting.

When the abbreviation sentence complementing unit 350 adds the morphemes constituting the discourse range to the first morphological information, the emotional state determining unit 381 ranks the emotional degree held by the user at a higher rank. Attaching steps can also be performed.

More specifically, the abbreviation sentence complementing section 350 selects from among the attributes (the subject case composed of the nominative case, the target case composed of the object case, etc.) constituting the first morpheme information extracted by the sentence structure analyzing unit 430. , Search for attributes that do not include morphemes. The abbreviation sentence complementing unit 350 adds, based on the retrieved attribute, a morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute. When the morpheme constituting the discourse range is added to the first morpheme information by the abbreviation sentence complementing unit 350, the emotion state determination unit 381 ranks the degree of emotion held by the user at a higher rank.

That is, in the abbreviation sentence complementing section 350, the input information consisting of the first morpheme information is replaced with the abbreviation sentence (when the discourse range is added to the first morpheme information, the first morpheme information before the morpheme is added) Means that the input information consisting of is abbreviated sentence), the user assumes that the input information belonging to the current discourse range belongs to the current discourse range. Since the input information is considered to have been input, the input information input at this stage is likely to be an abbreviated sentence omitting the morpheme consisting of the current discourse range (details are described in the first embodiment. (Refer to the section of “abbreviated sentence complementing section 350”).

When the abbreviation sentence complementing unit 350 determines that the first morpheme information is an abbreviation, the emotion state determination unit 381 sets the current discourse range to be substantially the same as the previous discourse range. The first morpheme information is determined to be a cohesion factor having relevance to the previous discourse range. In this case, if the unity factor is the unity factor (+), the emotion state determination unit 381 sets the degree of emotion to the user to be higher, and if the unity factor is the unity factor (-), It is possible to set a lower degree of emotion for the user.

[0320] Furthermore, the emotional state determination unit 381 can also perform a step of ranking the degree of emotion held by the user based on the search result by the reflexive determination unit 320 or the parrot return determination unit 330. This reflexive determination unit 320
Is a fixed form obtaining means for comparing the extracted first morphological information with each fixed form content and searching for reflective element information including the first morphological information from each fixed form content. Specifically, when the reflexive determination unit 320 searches for the fixed content that matches the first morpheme information from among the fixed content, the emotional state determination unit 381 ranks the user with a lower degree of emotion. Rank.

That is, the fact that the reflexive judging unit 320 has searched for the fixed content that matches the first morpheme information is an element (interruption factor) for interrupting the current topic by the first morpheme information. Therefore, the emotional state determination unit 381 determines that the first morpheme information is a negative emotional factor (interruption factor), and sets the emotional level to the user to be lower. (Fig. 26,
See FIG. 27). For example, if the user inputs “Good morning” (regular content) that is completely unrelated to the topic during the conversation, “Good morning” becomes a cause of interruption.

The parrot return judging section 330 uses the current first morphological information extracted by the morphological extracting section 410 as the past answer contents extracted by the morphological extracting section 410 (this past first morphological information is Element database 802
(Which is stored in the memory).

If the parrot return determining unit 330 determines that the current first morpheme information is included in the past answer contents, the emotional state determining unit 381 determines that the current first morphological information is included in the past answer contents. The user is parroting the past answer (the user is listening again to the answer), and the first morphological information is the interruption factor for interrupting the current topic Can be quasi-determined.

[0324] In this case, since the current first morphological information is a negative emotion factor (interruption factor), the emotion state determination unit 381 sets the degree of emotion held by the user to the lower one. (See FIGS. 26 and 27).

Further, the parrot return judging section 330 collates the current morpheme information extracted by the morpheme extraction section 410 with the past first morpheme information, and the current first morpheme information is included in the past first morpheme information. If the emotion state determination unit 381 determines that the current first morpheme information is included in the past first morpheme information,
It can be determined that the user has repeatedly input the same input information as the past input information. In this case, the emotion state determination unit 381 sets the degree of emotion to the user to be lower because the input information repeatedly input is the cause of the interruption.

If the reflexive judgment section 320 cannot search the fixed form contents that match the first morpheme information from each fixed form content, the emotional state judgment section 381 increases the degree of emotion held by the user. You may rank the rank. When the parrot return determining unit 330 cannot search for past first morphological information or past answer content that matches the current first morphological information, the emotional state determining unit 381 ranks the user with a higher degree of emotion. May be ranked.

[0327] Furthermore, when the input information specified by the utterance type determination unit 440 is a repulsive element such as a repulsive sentence, the emotional state determination unit 381 ranks the degree of emotion held by the user at a lower rank. Attaching steps can also be performed.

That is, the fact that the utterance type determination unit 440 specifies that the input type of the input information input by the user is composed of a repulsive element means that the user and the conversation control device 1 have opposition (oppose). ), The emotional state determination unit 381 determines that the first morphological information is an emotion minus factor (conflict factor), and determines the degree of emotion held by the user. It is set to the lower one (see FIGS. 26 and 27).

Next, if the topic search unit 360 does not retrieve the second morpheme information including the first morpheme information from the respective second morpheme information, the understanding state determination unit 382 determines that the input information Is difficult (harass) (factor of difficulty)
Is determined, and a step of ranking the understanding of the input information to a lower rank is performed (S211).

Specifically, the fact that the second morpheme information including the first morpheme information is not retrieved from the second morpheme information by the topic retrieval unit 360 is more difficult than the second morpheme information. Therefore, the understanding state determination unit 382 determines that the first morphological information is a cause of difficulty, and sets the understanding level of the input information to a lower level.

When the topic search unit 360 searches the second morpheme information for the second morpheme information including the first morpheme information, the understanding state determination unit 382 determines whether the input has been input by the user. It may be determined that the information is easy, and the understanding of the input information may be ranked higher. When the first morpheme information is a factor that causes a topic change to a cohesion factor, the understanding state determination unit 382 may rank the understanding level of the input information to a higher rank.

Next, when the input information specified by the utterance type judgment unit 440 is a repulsion element (oppose; also called an opposition factor) such as a repulsion sentence,
A step of ranking the degree of conflict with the user to a higher rank is performed (S212). Also, the conflict state determination unit 383
If the input information specified by the utterance type determination unit 440 is not a repulsive element such as a repellent sentence, a step of ranking the degree of conflict with the user to a lower rank is performed.

After that, the emotion state determination section 381, the understanding state determination section 382, or the conflict state determination section 383 outputs the determined emotion level, understanding level, and conflict level to the output section 600, and the output section receives the input. A step of displaying the degree of emotion, the degree of understanding, and the degree of conflict is performed (S213).

(Operation and Effect of Information Processing System and Information Processing Method) According to the invention according to the present embodiment, the reflexive determination unit 320, the parrot return determination unit 330, or the topic search unit 360 transmits the first morpheme information And the second morpheme information, and searches the second morpheme information for the second morpheme information including the first morpheme information.
81 can be ranked as the degree of emotions held by the user according to the magnitude of the frequency at which the second morphological information including the first morphological information is searched from among the pieces of second morphological information. Therefore, if the emotional state determination unit 381 can output (for example, a screen, a voice, or the like) the emotional level of the user with the ranked size, the conversation control device 1 uses It can be used to specify the degree of emotion held by the person.

Further, the developer who develops the information processing system can easily rank, for example, the degree of emotion based on the search history performed between the first morpheme information and each of the second morpheme information. Therefore, there is no need to develop a complicated algorithm or a neural network for causing the information processing system to calculate the degree of emotion held by the user, and as a result, the development time for developing the present system is reduced. be able to.

The discourse range determining unit 340 determines whether the first discourse range associated with the first morpheme information retrieved first and the second discourse range associated with the first morpheme information retrieved later. Emotional state determination unit 38
1 ranks the degree of emotion held by the user based on the result determined by the discourse range determination unit 340, so that the emotion state determination unit 381 compares the first morpheme information with each second morpheme information. It is possible to rank the degree of emotion held by the user more appropriately than to rank the degree of emotion held by the user based on the search result based on the search result.

[0337] For example, if the second discourse range has no relevance with the first discourse range, the emotional state determination unit 381 precedes the current input information (input by the user). Since it is possible to judge that the input information is completely different from the input information, the user can falsely determine that the input is inconsistent with respect to a certain topic. The degree of emotion to the person can be set low (bad direction).

On the other hand, if the second discourse range has relevance with the first discourse range, the emotional state determination unit 381 will add the current input information (input from the user) before Since it can be determined that the input information is related to the input information, the user can falsely determine that the input is consistent with a certain topic, and Can be set high (good direction).

[0339] Further, when the morpheme of the discourse range is added to each attribute constituting the first morpheme information by the abbreviation sentence complementing unit 350, the emotion state determination unit 381 omits the input information composed of the first morpheme information. It can be determined that the sentence is a sentence, and it can be determined that the user has input the input information on the premise of the previously searched discourse range.

For this reason, when the first morpheme information consisting of the input information is an abbreviation,
Since it can be determined that the input information has been input based on the previously searched discourse range, the user can quasi-determine that the input is consistent for a certain topic It is possible to set a high degree of emotion for the user (in a good direction).

Further, the reflexive judgment section 320 compares the first morpheme information with each of the fixed form contents, searches the fixed form contents for the fixed form contents including the first morpheme information, and determines the emotional state judgment section 381. Ranks the degree of emotion held by the user based on the search result.
For example, when a fixed content including the first morphological information is searched, the user can quasi-determine that the user has input the input information for interrupting a certain topic, The degree of emotion held by the user can be ranked lower, and as a result, the degree of emotion held by the user can be ranked more appropriately.

On the other hand, when the reflexive judgment section 320 cannot search for the fixed form content including the first morpheme information, the emotional state judgment section 381 sets the user's input for interrupting a certain topic. Since no information has been entered, it can be quasi-determined that the user is seriously answering the current topic, so that the degree of emotion held by the user can be ranked high. .

Further, the utterance type determination unit 440 specifies an input type indicating the type of input information, and if the specified input type is a repulsion element such as a repulsion sentence, the emotion state determination unit 381 uses The degree of emotion held by the user can be ranked at a lower rank, and the degree of emotion held by the user can be ranked more appropriately.

Further, if the topic search unit 360 does not search for the second morpheme information including the first morpheme information, the understanding state determination unit 382 stores the second morpheme information related to the first morpheme information in advance. This means that the input information consisting of the first morpheme information is more difficult to understand than the sentence consisting of the second morpheme information.

[0345] Therefore, the understanding state determination unit 382 cannot understand the input information consisting of the first morpheme information, and therefore, it is supposed that the user has input that the meaning is unknown for a certain topic. And the degree of understanding of the input information can be ranked to a lower rank.

If the understanding state determination unit 382 can retrieve the second morpheme information related to the first morpheme information by the topic search unit 360, the understanding state determination unit 382
Can pseudo-determine that the user has input meaningful information for a certain topic, and can rank the understanding level of the input information to a higher rank.

Further, the developer who develops the conversation control device 1 can hierarchically construct the second morpheme information and the like for retrieving the answer to the user's input information using a language such as DKML. Therefore, the conversation control device 1 can search the database for the second morpheme information and the like related to the first morpheme information based on the first morpheme information corresponding to the input information of the user through a hierarchical procedure. it can.

That is, the reflexive determination section 320, the parrot return determination section 330, the discourse range determination section 340, or the topic search section 360
Determines the hierarchy of the first morpheme information corresponding to the input information (for example, whether the second morpheme information stored in the database is in a higher concept or a lower concept), and determines the determined hierarchy. Based on the second morpheme information stored in advance, appropriate second morpheme information can be searched.

For this reason, the reflexive judgment section 320, the parrot return judgment section 330, the discourse range determination section 340 or the topic search section 3
60 is first morpheme information composed of user input information;
Since each "specific" second morpheme information belonging to a specific hierarchy and the first morpheme information need only be compared without having to match "all" second morpheme information stored in advance one by one. The second morpheme information that is similar to the one morpheme information can be searched in a short time, and the result can be ranked in a short time based on the search result, for example, the degree of emotion held by the user.

[0350] Further, the parrot return determination unit 330 compares the current first morpheme information with the past answer contents, and when it is determined that the current first morpheme information is included in the past answer contents, the emotional state is determined. The determination unit 381 can simulately determine that the user has returned a parrot to the past answer contents (that the user is listening again to the answer contents).

[0351] In this case, the emotion state determination unit 381 simulates that the user has input the input information for interrupting a certain topic because the user has performed a parrot on the past answer contents. It is possible to determine the degree of emotion held by the user at a lower rank, and as a result, the degree of emotion held by the user can be ranked more appropriately.

Finally, the parrot return judging section 330 compares 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 emotional state determination unit 381 can quasi-determine that the input information that the user has just input in the past has been re-input irrespective of the answer content from the conversation control device 1.

[0353] In this case, the emotional state determination unit 381 simply has to repeatedly input the same input information as the previous input information, irrespective of the answer content from the conversation control device 1. Because of this, the degree of emotion held by the user can be ranked at a lower rank, and as a result, the degree of emotion held by the user can be ranked more appropriately.

[Third Embodiment] (Basic Configuration of Access System) An embodiment of the present invention will be described with reference to the drawings. FIG. 31 is a schematic configuration diagram illustrating the access system according to the present embodiment.

As shown in the figure, the access system according to the present embodiment is provided from the Web servers 2a to 2c distributed on the communication network 1000 and from the Web servers 2a to 2c in response to a user request. Conversation control devices 1a to 1c for accessing Web pages to be executed.

FIG. 32 shows the internal configuration of the access system according to this embodiment. As shown in the figure, the conversation control device 1 included in the access system has almost the same internal structure as the conversation control device 1 in the second embodiment, but an access information search unit 390 is used instead of the answer sentence search unit 370. In that it has Except for this difference, the structure is the same as that of the first embodiment and the second embodiment, and the description of the structure other than the difference will be omitted.

In the first embodiment, a process in which the conversation control device 1 obtains an optimum answer sentence corresponding to input information based on input information from a user will be described. A description has been given of a process in which the ranking unit 380 in the 300 ranks the degree of emotion or the like held by the user based on input information from the user.

[0358] In the present embodiment, the conversation control device 1 displays the optimum Web page (site) corresponding to the input information based on the input information from the user, and is ranked by the ranking unit 380. Processing for displaying a Web page that more closely matches the user's request based on the rank size will be described. The specific description is as follows.

The access information search section 390 is an address obtaining means for obtaining access information associated with the second morpheme information based on the second morpheme information searched by the topic search section 360. The access information search unit 3
Reference numeral 90 is for acquiring access information associated with the second morpheme information based on the second morpheme information selected by the selection unit 362. This access information is
This is for accessing a Web page provided from the servers 2a to 2c. In the present embodiment, for example, U
RL (Uniform Resource Locator) and the like.

In the conversation database 500 (access information storage means), a plurality of pieces of second morphological information (topic titles) and access information are stored in association with each other in advance. In addition, a plurality of second morpheme information is associated with a discourse range indicating morphemes that are relevant to the input information that will be input from the user or the answer to the user, and each second morpheme information is The access information is associated with each of them, and the conversation database 500 is also a discourse storage unit that stores a plurality of discourse ranges in advance. In the present embodiment, the access information is described as being associated with all the second morpheme information. However, the present invention is not limited to this. Some are not associated with information.

As shown in FIG. 33, each of the discourse ranges A to D includes topics 1 to n (the ones belonging to the discourse range) meaning the range most closely related to the input information input by the user. ) Is associated with each of these topics 1 to
In n, a plurality of topic titles (second morpheme information) and a plurality of access information (URLs) are associated.

The URL may be, for example, an address for accessing company A's homepage, company information, financial information, employment information, product information, and the like in company A. As shown in the figure, the “discussion range” is positioned as a superordinate concept of “topic” in the present embodiment.

[0363] Specifically, the access information search unit 390 to which the search result signal is input from the topic search unit 360 is associated with the second morpheme information based on the second morpheme information corresponding to the input search result signal. Access information
It is acquired from the conversation database 500. For the search of the second morpheme information performed by the topic search unit 360, refer to the first embodiment.

The access information search unit 390 is also an address obtaining means for obtaining access information associated with the specific information based on the specific information searched by the topic search unit 360.

[0365] The topic search unit 360 includes the sentence structure analysis unit 4
The input type specified by the utterance type determination unit 440 and the specific information associated with the second morphological information are referred to based on the second morphological information retrieved in step 30, and the input type is determined from among the specific information. This is to search for specific information that matches. Here, the specific information is affirmative constituting the input information,
This is for specifying a type such as negative or repulsion.

A plurality of access information items are associated with the topic title (second morpheme information), and each of the access information items is specified for specifying the type of the input information, such as affirmative, negative, or negative. Information is associated with each of them, and the conversation database 500 stores a plurality of topic titles in advance.

As shown in FIG. 34, for example, the discourse range (product A), subordinate concepts (amusement, pachinko,
Pachislot, games), synonyms (new products, new models, products,
If the subject title 1 belongs to this discourse range, for example, (product; *; exists) {is there any product? }, The URL 1-1 associated with the topic title 1 is URL-1-1-1 to 1-1-5 associated with each piece of specific information in advance.

More specifically, the second morpheme information (product; *; present) retrieved from the topic retrieval unit 360 and the “utterance sentence type” (DA) (input type; For details, the access information search unit 390 to which the “utterance type determination unit 440” in the first embodiment is input) is associated with the input second morpheme information (product; *; exists). The specific information obtained is collated with the input "type of utterance sentence" (DA), and specific information (DA) matching the "type of utterance sentence" (DA) is searched from among the specific information. , Searched specific information (DA)
34. Select the URL 1-1-1 associated with.
reference).

[0369] Also, the access information search unit 390, based on the second morpheme information searched by the topic search unit 360,
It is also a rank obtaining unit that checks each priority associated with the second morpheme information with the rank ranked by the ranking unit 380, and obtains a priority that matches the rank from each priority. Further, the access information search unit 39
0 is an access information acquisition unit that acquires access information associated with the priority based on the acquired priority.

[0370] A plurality of pieces of access information are associated with the second morpheme information, and each access information has a priority order to be selected as access information corresponding to the input information input by the user. The conversation database 500 stores a plurality of pieces of the second morpheme information in advance.

As shown in FIG. 35, for example, the discourse range (product A), subordinate concepts (amusement, pachinko,
Pachislot, games), synonyms (new products, new models, products,
If the subject title 1 belongs to this discourse range, for example, (product; *; exists) {is there any product? }, URL1-1-1 to 1-1-5 are associated in advance with the respective priorities as the URL1-1 associated with the topic title 1.

More specifically, the second morpheme information (product; *; exists) retrieved from the topic retrieval unit 360 and the ranking unit 3
The access information search unit 390 to which the rank (3) specified from 80 has been input is assigned to each of the priorities associated with the input second morphological information (product; *; there is) and the size of the specified rank. (3), and the priority (3) that matches the size of the rank is specified from among the priorities, and the URL1-1 associated with the specified priority is determined.
Select 3. Access information search unit 3 selecting a URL
90 outputs the selected URL to the communication unit 800.

Communication unit 800 includes access information search unit 39
Access means for accessing a Web page corresponding to the access information based on the access information acquired at 0. Specifically, the communication unit 800 to which the URL selected from the access information search unit 390 has been input is
The web page corresponding to L is accessed, a file of the accessed web page is obtained, and the obtained file is output to the output unit 600. This file is, for example, HT
Examples include sentences written in ML (Hyper Text Markup Language).

The second morpheme information including the first morpheme information (including the discourse range, the fixed content, the answer content, etc.
“Abbreviated as“ second morpheme information, etc. ”) means not only the case where the second morpheme information etc. directly includes the first morpheme information, but also the case where the second morpheme information etc. includes at least one morpheme constituting the first morpheme information. It also means the case.

(Operation and Effect of Access System)
According to the invention according to the present embodiment, the topic search unit 360 determines the morpheme constituting the first morpheme information from among the second morpheme information (the morpheme constituting the input information input by the user). The access information search unit 390 acquires the access information associated with the second morpheme information based on the retrieved second morpheme information, and the communication unit 800 acquires the acquired second morpheme information. Based on the access information, it is possible to access a Web page corresponding to the access information.

As a result, the conversation control device 1 stores the semantic space constructed by each morpheme (the semantic content grasped from the character string composed of each morpheme) based on each morpheme constituting the input information from the user. It is possible to acquire the access information associated in advance based on the information, and to access the Web page associated with the acquired access information, so that the user can input a keyword related to the site requested by the user one by one. Instead, necessary information can be automatically acquired only by having a conversation with the conversation control device 1. Further, the communication unit 800 does not simply access the address of the Web page associated with the keyword based on a predetermined keyword input by the user, but the user's utterance content (input information).
Can be accessed on the basis of each morpheme that constitutes the morpheme, and as a result, the address of a specific Web page that takes into account the semantic content requested by the user is accessed. The output unit 60 can output a site that matches the user's request.
0 can be displayed.

[0377] Further, the utterance type determination unit 440 specifies the type of the input information, such as affirmative, negative, or repulsion, as the input type based on the character string of the input information input by the user, and performs access. The information search unit 390 collates the specified input type with each piece of specific information associated with the searched second morphological information, and searches each piece of specific information for specific information that matches the input type. The access information associated with the specific information can be obtained based on the specified information.

Thus, access information search section 390
Is the input type that constitutes the content of the user's utterance, for example,
Access information that matches the input type from multiple types of access information based on what the user simply gave an opinion, what the user had, and what the user described the locational element As a result, the communication unit 800 can cause the output unit 600 to display a site that matches the user's request.

Also, the topic search unit 360 checks the extracted first morpheme information against each second morpheme information, and for each second morpheme information, the first morpheme information occupies the second morpheme information. Calculate the ratio, according to the magnitude of each ratio calculated for each second morpheme information, from among each second morpheme information,
One second morpheme information is selected, and the access information search unit 39
0 can obtain the access information associated with the second morpheme information based on the selected second morpheme information.

[0380] Thus, the topic search unit 360 divides the second morpheme information having a large proportion of the first morpheme information (elements constituting the input information of the user) into the second morpheme information, for example, into each second morpheme information. If it can be obtained from the group, it is possible to more accurately obtain the second morphological information following the semantic space composed of the first morphological information,
As a result, the access information search unit 390 can more accurately acquire access information that matches the user's request based on the second morphological information.

Further, the morpheme extraction unit 410 collates the morpheme extracted from the character string with each set group, selects one set group including the morpheme from each set group, and associates the set group with the selected set group. Since the extracted element information can be extracted as the first morpheme information, the morpheme extraction unit 410 determines that the “present” is, for example, a present, a gift, a gift, a celebration when the first morpheme information is “present”. Therefore, element information associated with the set group, for example, “gift” can be extracted as the first morpheme information.

Thus, the developer who develops the conversation control device 1 compares the second morpheme information corresponding to the meaning grasped from the mutually similar first morpheme information and the answer content related to the second morpheme information with each other. It is no longer necessary to create one for each first morpheme information to be executed.
Since the amount of data stored in the conversation control device 1 can be reduced, and the load on the conversation control device 1 is reduced, the access information necessary for the user can be quickly acquired.

Also, the topic search unit 360 determines each morpheme of the first morpheme information belonging to each of the classified attributes (subject, subject, etc.) and each of the second morpheme information belonging to each of the attributes stored in advance. The morpheme is compared with each attribute, and the second morpheme information including the morpheme of the first morpheme information in at least one attribute can be searched from the second morpheme information. Based on the order of words constituting the input information, the second morpheme information closest to the word order can be obtained.

That is, the topic search unit 360 inputs input information that is difficult to identify, for example, “a person bites a dog” and “a dog bites a person”.
Can be identified, the second morpheme information matched by the first morpheme information constituting the input information can be obtained, and as a result, the access information search unit 390 can be obtained.
Can acquire access information that matches the user's request based on the acquired second morpheme information.

Also, the topic search unit 360 checks each second morpheme information belonging to the “discussion range” with the first morpheme information to search for second morpheme information that is similar to the first morpheme information (pinpoint It is not necessary to match “all” of the second morpheme information with the first morpheme information. Therefore, until the second morpheme information that is similar to the first morpheme information is searched. Time can be shortened.

As a result, the topic search unit 360 searches for the second morpheme information similar to the first morpheme information in a short time, so that the access information search unit 390 is associated with the searched second morpheme information. The communication unit 800 can quickly acquire a site that matches the user's request based on the acquired access information.

Also, the ranking unit 380 ranks the second morpheme information including the first morpheme information from the respective second morpheme information in accordance with the frequency of the search, and the access information search unit 390 , Based on the retrieved second morphological information, collates each priority associated with the second morphological information with the ranked rank, and from each priority, the priority that matches the rank And the access information associated with the priority can be obtained based on the obtained priority.

As a result, if the rank ranked by the ranking unit 380 with reference to the above search result (search history) is the degree of emotion held by the user, the access information search unit 390 will The conversation control device 1 can specify the degree of emotion of the user, and automatically specify the specific Web page based on the specified degree of emotion. It can be used as a device for displaying.

[0389] If a Web page of interest to the user is provided according to the priority order, the access information search unit 390 can refer to the search result when the rank ranked is large. Can display a Web page of particular interest to the user.

As a result, for example, if the user does not input the input information including the first morpheme information that can be searched by the topic search unit 390 or the like, the topic search unit 390 or the like matches the first morpheme information. The second morpheme information cannot be searched, and the ranking unit 380 sets the rank to a lower rank. As a result, it becomes impossible to acquire information that is of interest to the user as described above.

[0391] For this reason, the user must input input information that does not set the rank ranked by the ranking section 380 low in order to obtain information of interest to the user. Information about a specific Web page can be acquired while enjoying the feeling of having a conversation with the user with care.

Also, since the developer who develops the conversation control device 1 can specify the size ranked by the conversation control device 1 as the emotion degree or the like, the developer can express the emotion degree or the like held by the user in the conversation. There is no need to develop a complicated algorithm or a neural network for calculation by the control device 1, and as a result, the development time for developing a device that achieves the present invention can be reduced.

The stored information including the second morpheme information and the access information includes a topic desired by the user, a degree of emotion or a statement held by the user in a certain matter, a positive sentence, a question sentence, and a repulsive sentence. And the like are preferably described in a language for hierarchically storing them in a database in association with morphemes according to their semantic contents. This language is D
It is preferably composed of KML (Discourse Knowledge Markup Language). Accordingly, the developer who develops the conversation control device 1 can hierarchically construct the second morphological information and the like for searching for the answer content to the user's input information in the database using the language. The topic search unit 360 and the like can search the database for second morpheme information and the like related to the first morpheme information based on the first morpheme information corresponding to the input information of the user through a hierarchical procedure. .

That is, the topic search unit 360 determines whether the hierarchical level of the first morphological information corresponding to the input information (for example, whether the hierarchical level is higher or lower than the second morphological information stored in the database). ), And appropriate second morphological information can be retrieved from each of the second morphological information stored in advance based on the determined hierarchy.

For this reason, the topic search unit 360 does not collate the first morpheme information consisting of the input information of the user with “all” of the second morpheme information stored in advance, without having to collate each one. Since the second morpheme information of "specific" belonging to the first morpheme information may be compared with the first morpheme information, the second morpheme information similar to the first morpheme information can be searched in a short time. 800 is based on the access information associated with the retrieved second morpheme information,
A Web page corresponding to the access information can be quickly displayed.

[0396]

As described above, according to the present invention,
The conversation control device analyzes semantic contents that can be grasped from the utterance contents based on the utterance contents uttered by the user, obtains access information related to the analyzed semantic contents, and obtains the obtained access information. , It is possible to access a Web page that matches the user's request.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of a conversation control device according to a first embodiment.

FIG. 2 is a block diagram illustrating an internal structure of a conversation control unit and a sentence analysis unit according to the first embodiment.

FIG. 3 is a diagram showing contents of each morpheme extracted by a morpheme extraction unit in the first embodiment.

FIG. 4 is a diagram showing the content of each phrase extracted by a phrase analysis unit in the first embodiment.

FIG. 5 is a diagram illustrating the content of “case” specified by a sentence structure analysis unit in the first embodiment.

FIG. 6 is a diagram illustrating “utterance sentence type” specified by an utterance type determination unit in the first embodiment.

FIG. 7 is a diagram showing the contents of each dictionary stored in the utterance type database according to the first embodiment.

FIG. 8 is a diagram showing the contents of a hierarchical structure built inside a conversation database in the first embodiment.

FIG. 9 is a diagram showing a detailed relationship of a hierarchical structure constructed inside a conversation database in the first embodiment.

FIG. 10 is a diagram showing the contents of a “topic title” constructed inside a conversation database in the first embodiment.

FIG. 11 is a diagram showing the contents of “answer sentence type” associated with “topic title” constructed inside the conversation database in the first embodiment.

FIG. 12 “Topic title” belonging to “Discourse range” constructed inside the conversation database in the first embodiment.
FIG. 7 is a diagram showing the contents of a "answer sentence".

FIG. 13 is a diagram illustrating contents of reflection element information stored in a reflection element database according to the first embodiment.

FIG. 14 is a diagram showing parrot return elements stored in a parrot return element database and the contents of morphemes of the parrot return elements 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 arranged by a morpheme extraction unit in a first modified example.

FIG. 17 is a diagram illustrating an internal configuration of a topic search unit according to a second modification.

FIG. 18 is a diagram showing a case where the ratio calculation unit in the second modification is “case configuration”
FIG. 7 is a diagram showing a state in which each morpheme belonging to and is compared with each “topic title” for each “topic title”.

FIG. 19 is a diagram illustrating a case where the ratio calculation unit according to the second modification example is “each component”.
FIG. 11 is a diagram showing a state in which morphemes belonging to a “topic title” are collated with morphemes belonging to a “topic title” for each “case”.

FIG. 20 is a diagram illustrating a schematic configuration of a conversation control system according to a third modification.

FIG. 21 is a diagram illustrating an internal configuration of a topic search unit in a sixth modification.

FIG. 22 is a diagram illustrating a manner in which a topic search unit in the sixth modification example compares first morpheme information with second morpheme information or an answer sentence.

FIG. 23 is a diagram illustrating a recording medium that stores a program according to the present embodiment.

FIG. 24 is a block diagram illustrating an internal structure of a conversation control unit and a sentence analysis unit according to the second embodiment.

FIG. 25 is a diagram illustrating the contents of a “discussion event table” stored in a discourse event database according to the second embodiment.

FIG. 26 is a diagram illustrating a relationship between “topic change” and “cohesion factor” in the second embodiment.

FIG. 27 is a diagram illustrating a rank of an emotion degree determined by an emotion state determination unit according to the second embodiment.

FIG. 28 is a diagram illustrating a rank of an understanding level determined by an understanding state determination unit according to the second embodiment.

FIG. 29 is a diagram illustrating a rank of a degree of conflict determined by a conflict state determination unit according to the second embodiment.

FIG. 30 is a flowchart showing a procedure of a ranking method in the second embodiment.

FIG. 31 is a diagram illustrating a schematic configuration of an access system according to a third embodiment.

FIG. 32 is a diagram illustrating an internal structure of a conversation control device and a sentence analysis unit according to the third embodiment.

FIG. 33 is a diagram showing a discourse range group stored in a conversation database according to the third embodiment.

FIG. 34 is a diagram showing the contents of a “topic title” constructed inside a conversation database in the third embodiment (part 1).

FIG. 35 is a diagram showing the contents of a “topic title” constructed inside a conversation database in the third embodiment (part 2).

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Conversation control device, 100 ... Input part, 200 ... Speech recognition part, 300 ... Conversation control part, 310 ... Management part, 320 ... Reflective determination part, 321 ... Rate calculation part, 322 ... Selection part, 3
30: parrot return determination unit, 340: discourse range determination unit, 35
0: abbreviated sentence complementing section, 360: topic search section, 361: ratio calculating section, 362: selecting section, 363: deleting section, 364: discourse adding section, 370: answer sentence searching section, 380: ranking section, 381 ... Emotion state determination unit, 382: Understanding state determination unit, 383: Conflict state determination unit, 390: Access information search unit, 400: Sentence analysis unit, 410: Morphological extraction unit, 42
0: phrase analysis unit, 430: sentence structure analysis unit, 440: utterance type determination unit, 450: morphological database, 460: utterance type database, 500: conversation database, 60
0: output unit, 700: speech recognition dictionary storage unit, 800: communication unit, 801: reflection element database, 802: parrot return element database, 900: communication unit, 1000: communication network, 1100: hard disk, 1200 ...
Floppy (registered trademark) disk, 1300: compact disk, 1400: IC chip, 1500: cassette tape

   ────────────────────────────────────────────────── ─── Continuation of front page    (72) Inventor Hiroshi Katsukura             3-22-9 Takanawa, Minato-ku, Tokyo Arze Takanawa             Le (72) Inventor Kazuo Okada             Ariake, 3-1, Ariake, Koto-ku, Tokyo             Frontier Building (72) Inventor Jun Fujimoto             Ariake, 3-1, Ariake, Koto-ku, Tokyo             Frontier Building F term (reference) 5B075 ND03 PP07 PP24 PR08

Claims (16)

[Claims] 1. A W system distributed on a communication network.
An access system for accessing a Web page provided from an web server, comprising: a character recognition unit that specifies a character string indicating the input information based on input information input by a user; Based on the said character string,
Morpheme extraction means for extracting at least one morpheme constituting the minimum unit of the character string as first morpheme information; second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof; Access information storage means for storing a plurality of pieces of access information for accessing the web page provided from a web server in association with each other in advance; and the first morpheme information extracted by the morpheme extraction means and the respective second information First search means for comparing the morpheme information with the morpheme information and searching the second morpheme information for the second morpheme information including the morpheme constituting the first morpheme information; Address acquisition means for acquiring the access information associated with the second morpheme information based on the obtained second morpheme information; Access means for accessing the Web page corresponding to the access information based on the access information acquired by the acquisition means. 2. The access system according to claim 1, wherein a plurality of the access information are associated with the second morphological information, and each of the access information is affirmative or negative constituting the input information. Or, specific information for specifying the type of repulsion or the like is stored in advance in the access information storage means in association with each other, based on the character string specified by the character recognition means,
Type identification means for identifying the type of input information such as affirmation, negation or repulsion as an input type, and the input type identified based on the second morpheme information retrieved by the first retrieval means A second search unit that checks the specific information associated with the second morphological information with the specific information and searches the specific information for the specific information that matches the input type; Means for acquiring the access information associated with the specific information based on the specific information retrieved by the second retrieval means. 3. The access system according to claim 1, wherein the first search unit collates the first morpheme information extracted by the morpheme extraction unit with each of the second morpheme information, For each second morpheme information, calculating means for calculating the ratio of the first morpheme information to the second morpheme information, and the calculation means calculates the ratio of each of the ratios calculated for each of the second morpheme information. Selecting means for selecting one piece of the second morphological information from among the pieces of the second morphological information according to the size, wherein the address obtaining means is the second morphological element selected by the selecting means An access system, which acquires the access information associated with the second morpheme information based on information. 4. The access system according to claim 1, further comprising: an element storage unit configured to store a plurality of pieces of element information indicating the entire set group including a plurality of morpheme sets in advance in association with the set group. The morpheme extraction unit compares the morpheme extracted from the character string with each of the set groups, selects the set group including the morpheme from the set groups, and selects the selected set group. An access system for extracting the element information associated with the information as the first morphological information. 5. The access system according to claim 1, wherein each of the morphemes constituting the second morpheme information is classified according to attributes such as a subject matter composed of a nominative case and a target case composed of a purpose case. A classification unit pre-stored in an access information storage unit and classifying each morpheme of the first morpheme information extracted by the morpheme extraction unit into attributes such as the subject case and the subject case; Means, for each attribute, each morpheme of the first morpheme information belonging to each attribute classified by the classification means, and each morpheme of each second morpheme information belonging to each attribute stored in advance An access system for collating, and searching the second morpheme information, wherein the second morpheme information includes the morpheme of the first morpheme information in at least one of the attributes from the second morpheme information. 6. The access system according to claim 1, wherein a discourse range indicating the morpheme relevant to the input information to be input from the user or an answer to the user. A plurality of the second morpheme information are associated with each other, the access information is associated with each of the second morpheme information, and a discourse storage unit that stores a plurality of the discourse ranges in advance; A discourse search means for comparing the extracted first morpheme information with each of the discourse ranges, and searching for the discourse range including the first morpheme information from each of the discourse ranges; The second morpheme information associated with the identified discourse range is compared with the first morpheme information extracted by the morpheme extraction means, and the first morpheme information is extracted from the second morpheme information. First retrieval means for retrieving the second morpheme information including elementary information, wherein the address acquisition means determines the second morpheme information based on the second morpheme information retrieved by the first retrieval means. Acquiring an associated access information. 7. The access system according to claim 1, wherein a plurality of the access information are associated with the second morphological information, and each of the access information is associated with input information input by a user. A priority order to be selected as the corresponding access information is associated with each of the access information storage means for storing a plurality of the second morphological information in advance, and a search result searched by the first search means A ranking unit that ranks the second morpheme information including the first morpheme information based on the magnitude of the frequency of being searched, based on the second morpheme information, Based on the retrieved second morphological information, each priority associated with the second morphological information is compared with the rank ranked by the ranking means, and each priority is compared. And a priority obtaining unit that obtains the priority that matches the rank from among the above. The address obtaining unit is associated with the priority based on the priority obtained by the priority obtaining unit. Obtaining the access information described above. 8. The access system according to claim 1, wherein the second morpheme information stored in the access storage means and the storage information including the access information are a topic or a matter requested by the user. Described in a language for hierarchically storing in the database the degree of emotion or the type of statement, affirmative sentence, question sentence, repellent sentence held by the user in association with the morpheme according to the semantic content. An access system, characterized in that: 9. W distributed on a communication network
An access control method for accessing a Web page provided from an web server, comprising: a step of specifying a character string indicating the input information based on input information input by a user; Extracting at least one morpheme constituting the minimum unit of the character string as the first morpheme information based on the first morpheme information; one character, a plurality of character strings, or second morpheme information indicating a morpheme composed of a combination thereof; Storing a plurality of pieces of access information for accessing the web page provided from the web server in association with each other in advance; and comparing the extracted first morpheme information with each of the second morpheme information; Searching the second morpheme information from the respective second morpheme information, the second morpheme information including the morpheme constituting the first morpheme information Acquiring the access information associated with the second morpheme information based on the retrieved second morpheme information; and the web page corresponding to the access information based on the acquired access information. Access to the access control method. 10. The access control method according to claim 9, wherein the second morpheme information is associated with a plurality of the pieces of access information, and each of the pieces of the access information is affirmative constituting the input information. A step of storing a plurality of pieces of the second morpheme information in advance in association with specific information for specifying types such as denial or repulsion, and affirming the input information based on the specified character string; Specifying a type such as negative or repulsion as an input type; and collating the specified input type with the specific information associated with the second morphological information based on the retrieved second morphological information. Searching the specific information from the specific information for the specific information that matches the input type; and searching for the specific information associated with the specific information based on the searched specific information. Access control method characterized by a step of acquiring the access information. 11. The access control method according to claim 9, wherein the extracted first morpheme information and the second morpheme information are collated, and the second morpheme information is compared with the second morpheme information. Calculating the ratio of the first morpheme information to the morpheme information, and according to the magnitude of each ratio calculated for each of the second morpheme information, from among the respective second morpheme information, Selecting one of the second morpheme information; and obtaining the access information associated with the second morpheme information based on the selected second morpheme information. Control method. 12. The access control method according to claim 9, wherein a plurality of pieces of element information indicating an entire set group including a plurality of morpheme sets are stored in advance in association with the set group. The morpheme extracted from the character string is compared with the respective set groups, the set group including the morpheme is selected from the respective set groups, and the element information associated with the selected set group is selected. Extracting as first morpheme information. 13. The access control method according to claim 9, wherein each of the morphemes constituting the second morpheme information is classified by attributes such as a subject matter composed of a nominative case and a target case composed of a purpose case. Classifying each morpheme of the extracted first morpheme information into attributes such as the subject matter and the target case, which are stored in advance; and each of the first morpheme information belonging to each of the classified attributes. A morpheme is compared with each morpheme of the second morpheme information belonging to each attribute stored in advance for each attribute, and among the second morpheme information, at least one of the attributes is assigned to the first attribute. Retrieving the second morpheme information including the morpheme of the morpheme information. 14. The access control method according to claim 9, wherein the discourse range indicating the morpheme relevant to the input information to be input from the user or the answer to the user. A plurality of the second morpheme information are associated with each other, the access information is associated with each of the second morpheme information, and a plurality of the discourse ranges are stored in advance; Comparing the morpheme information with each of the discourse ranges, and searching for the discourse range including the first morpheme information from each of the discourse ranges; and A step of collating the two morpheme information with the first morpheme information extracted by the morpheme extraction means, and searching for the second morpheme information including the first morpheme information from each of the second morpheme information. And a step of acquiring the access information associated with the second morphological information based on the retrieved second morphological information. 15. The access control method according to claim 9, wherein a plurality of pieces of the access information are associated with the second morphological information, and each piece of the access information is input information input by a user. A priority order for being selected as the access information corresponding to the second morpheme information, a step of storing a plurality of the second morpheme information in advance, and the second morpheme information based on a retrieved search result. Ranking the second morpheme information including the first morpheme information from the information in accordance with the magnitude of the frequency of the search; and based on the searched second morpheme information, the second morpheme information A process of comparing each priority associated with information with the rank ranked by the ranking means, and acquiring, from each priority, the priority that matches the rank. An access control method, comprising: a step; and acquiring the access information associated with the priority based on the acquired priority. 16. The access control method according to claim 9, wherein the second morpheme information stored in the access storage means and the storage information including the access information are topics requested by the user, certain matters. Described in a language for hierarchically storing in the database hierarchically the degree of emotion or a sentence, an affirmative sentence, a question sentence, or a repellent sentence held by the user in association with the morpheme according to the semantic content. An access control method characterized by: