JP2003345604A - Language computer, language processing method, and program - Google Patents

Abstract

(57) [Summary] A language computer using a semiotic base is provided. SOLUTION: A context base 111 of a semiotic base 102 of a language computer 101 stores situation description knowledge explaining a feature of a situation, knowledge in a situation explaining knowledge used in a certain situation, and matters related to a user. Describe and store the user profile to be described in natural language,
The situation-dependent dictionary 116 holds situation characteristics, vocabulary information and meaning information of a language under a certain situation, and the input receiving unit 1
03 receives an utterance input from the user, the situation identification unit 104 identifies the utterance situation with reference to the semiotic base 102, and the text generation unit 105 associates the utterance situation with the identified situation. , By referring to the situation description knowledge, the knowledge in the situation, the user profile, and the vocabulary information and semantic information of the language stored in the context-dependent dictionary, which are described and stored in the context base 111 in a natural language. Generate text.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a language computer, a language processing method, and a program for realizing these on a computer.

[0002]

2. Description of the Related Art The inventors have traditionally used natural languages, in particular,
We are conducting research on methods of operating computers using languages that ordinary users who are not familiar with computers use in daily conversations (also called "daily languages").
Japanese Patent Application No. 2001-334 discloses a basic technology for this purpose.
No. 64 proposes a language computer using a semiotic base.

[0003] A semiotic base is a structured structure of the meaning of everyday language, as disclosed in the application. This application discloses a technology that uses a semiotic base to interpret and understand a conversation sentenced by a general user and performs processing according to the contents of the conversational sentence.

[0004]

[Problems to be solved by the invention]
In the processing of everyday language, it is necessary to make an inquiry to the general user if necessary according to the situation where the uttering general user is located, and to interpret and understand the processing required by the general user through dialogue. However, it is desirable to use natural language and everyday language for inquiries to general users.

The present invention has been made to solve the above problems, and has as its object to provide a language computer using a semiotic base, a language processing method, and a program for realizing these.

[0006]

In order to achieve the above object, the following invention is disclosed in accordance with the principle of the present invention.

[0007] A language computer according to a first aspect of the present invention includes a semiotic base, an input receiving unit, a situation identifying unit, and a text generating unit, and is configured as follows.

That is, the semiotic base is composed of a context base, a semantic base, a lexical grammar base,
It has a corpus base and a general dictionary.

[0009] Here, the context base describes situation description knowledge explaining the features of the situation, knowledge in the situation explaining knowledge used under a given situation, and matters related to the user using the language computer. The user profile to be written is described in a natural language and stored.

On the other hand, the semantic base holds the semantic features of the language and the corresponding semantic roles.

Further, the lexical grammar base holds the lexical grammatical features of the language.

The corpus base holds a language text as an example of language exchange, together with a semantic feature, a situation feature, and a lexical grammar feature.

On the other hand, the general-purpose dictionary holds vocabulary information and semantic information of a language.

Further, the situation-dependent dictionary holds situation characteristics and vocabulary information and meaning information of a language under a given situation.

On the other hand, the input receiving unit receives an utterance input from the user.

Further, the situation identification unit identifies the situation of the utterance input from the user who has received the input, with reference to the semiotic base.

The text generator is stored in a semiotic base including the situation description knowledge, the in-situation knowledge, and the user profile which are described in a natural language and stored in the context base in association with the identified situation. With reference to the received information, a text of a response to an utterance input from the user who has received the input is generated.

Further, in the language computer of the present invention, the text of the response generated by the text generation unit is a sentence included in the description of the situation description knowledge, the knowledge in the situation, and the user profile in the natural language stored in the context base. It can be configured to include.

In the language computer of the present invention, the semiotic base further includes a context-dependent dictionary, and can be configured as follows.

That is, the situation-dependent dictionary holds situation features and vocabulary information and meaning information of a language under a given situation.

On the other hand, the text generation unit generates the text of the response by further referring to the vocabulary information and the semantic information of the language under the identified situation stored in the context-dependent dictionary.

A language processing method according to a second aspect of the present invention includes a language processing, an input receiving step, a situation identification step, and a text generation step, and is the same as a semiotic used in the above-described language computer. Using the base, the configuration is as follows.

That is, in the input accepting step, an utterance input from a user whose items are stored in the context base is accepted.

On the other hand, in the situation identification step, the situation of the utterance input from the user who has accepted the input is determined.
Identify with reference to semiotic base.

Further, in the text generating step, the situation description knowledge, the knowledge in the situation, which is described in a natural language and stored in the context base in association with the identified situation,
With reference to the information stored in the semiotic base including the user profile, a text of a response to an utterance input from the user who has received the input is generated.

A program according to another aspect of the present invention causes a computer to function as the above-described language computer, or causes the computer to execute the above-described language processing method.

The program can be distributed and sold independently of a computer by recording the program on an information recording medium, and can be directly distributed and sold to a user via a computer communication network such as the Internet.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below. The embodiments described below are for explanation, and do not limit the scope of the present invention. Therefore, those skilled in the art can adopt embodiments in which each of these elements or all elements are replaced with equivalents, and these embodiments are also included in the scope of the present invention.

(First Embodiment) FIG. 1 is a schematic diagram showing a schematic configuration of a language computer according to one embodiment of the present invention. FIG. 2 is a flowchart showing a processing flow of a language processing method executed by the language computer. Hereinafter, description will be made with reference to these drawings.

The language computer 101 includes a semiotic base 102, an input receiving unit 103, and a situation identifying unit 1
04 and a text generation unit 105.

FIG. 3 is an explanatory diagram showing a schematic configuration of the semiotic base 102. Hereinafter, description will be made with reference to this figure.

The semiotic base 102 has a context base 111, a meaning base 112, a vocabulary grammar base 113, a corpus base 114, a general-purpose dictionary (generalized electronic dictionary) 115, and a context-dependent dictionary 116.

The context base 111 includes situation description knowledge explaining the features of the situation, knowledge in the situation explaining knowledge used under a given situation, and a user profile explaining items related to the user using the language computer. Is described and stored in natural language. As described above, one of the features of the present embodiment is that knowledge is described and stored in the context base 111 in a natural language or a daily language.

On the other hand, the semantic base 112 holds the semantic features of the language and the corresponding semantic roles.

Further, the lexical grammar base 113 holds the lexical grammatical features of the language.

The corpus base 114 holds a language text as an example of language exchange, together with semantic features, situational features, and lexical grammar features.

On the other hand, the general-purpose dictionary 115 holds vocabulary information and meaning information of a language.

Further, the situation-dependent dictionary 116 holds situation features and vocabulary information and meaning information of a language under a given situation. As described above, one of the features of the present embodiment is that information under the situation given to the situation-dependent dictionary 116 is held.

When the language processing method is started, first, the input receiving unit 103 receives an utterance input from the user (step S201).

Next, the situation identifying unit 104 identifies the situation of the utterance input from the user who has received the input by referring to the semiotic base 102 (step S202).

Then, the text generator 105 associates the context base 11 with the identified situation.
Reference is made to the information stored in the semiotic base 102 including the situation description knowledge, the knowledge in the situation, and the user profile which are described and stored in natural language, and the utterance input from the user who received the input is described. A response text is generated (step S203).

Here, in the present embodiment, the text of the response generated by the text generation unit 105 is a sentence included in the description of the situation description knowledge, the knowledge in the situation, and the user profile stored in the context base 111 in the natural language. Include. This is performed by using the knowledge stored in the context base 111 and described in the natural language and the everyday language.

In the present embodiment, the text generator 1
05 further refers to the vocabulary information and semantic information of the language under the identified situation held in the situation-dependent dictionary 116 to appropriately and quickly understand the utterance and progress of the dialogue according to the situation. To generate the text of the response.

Hereinafter, each element in these embodiments will be described in detail.

(Context Base) Hereinafter, the context base 111 of the present embodiment will be described in detail with reference to FIG. For easy understanding, matters disclosed in Japanese Patent Application No. 2001-33464 are not described as appropriate (the same applies hereinafter).

First, the context base 111 can be classified into the following three types.・ Context-based. Handles the situation and related knowledge.・ Text system. Handles the progress of the text.・ Interaction system. Handles the progress of interactions that take place in real time.

Then, the context system of the context base 111 stores the situation and the knowledge about the given situation.

On the other hand, each component of the semiotic base 102 has the following three roles. -Roles related to idea constructive functions. -Roles related to interpersonal functions. -Roles related to text-forming functions.

The knowledge stored in the context base 111 is also classified according to these roles.

First, in the context system, knowledge of the situation itself is a situation type corresponding to all three roles,
It is classified into the field of activity (Field) corresponding to the idea-constitutive function, the role relationship (Tenor) corresponding to the interpersonal function, and the communication mode (Mode) corresponding to the text-forming function.

The knowledge about a given situation can be classified into the following three types.・ Knowledge of situation description. The characteristics of the situation will be described.・ Knowledge within the situation. Explain the knowledge used in a given situation. -User profile. The basic information of the user, the knowledge level about the computer, the record of the past work using the computer, and other items necessary for the user to operate the computer will be described.

These three kinds of knowledge are described in a natural language, and are further provided with semiotic-based tags.

For example, records relating to past work using a computer include the contents of utterances in natural and everyday languages used by the user in past work.

The concepts of knowledge used in a given situation are categorized into verb concepts and noun concepts, and are connected through concept frames and concept connectors. In addition, "events corresponding to the result of connecting verb concepts and noun concepts" are also connected to each other to form a concept network.

On the other hand, knowledge such as genre structure, stage, and move is classified into the text system. The genre structure describes the text scene and the development of the utterance in it.

The interaction system includes discourse management knowledge and utterance management module, and categorizes dialog management knowledge, utterance contents, utterance function, and subject structure.

The discourse management knowledge is used in the text planning of the speaker and the listener. The utterance management module recognizes and manages the current interaction state. These work together.

FIG. 4 is a schematic diagram showing an outline of situation description knowledge. The situation description knowledge shown in the figure is knowledge about a task, and is related to "changing a character font".

This knowledge includes information such as an active area, a role relationship, and a communication style according to the above structure.

In the “Overall Wordings”, natural and everyday language expressions such as “the user creates sentences on a computer” and “the data is displayed on a display” are represented by “<Sphere of action 1>”, </ Sphere of action1
> ”,“ <Sphere of action4> ”,“ </ Sphere of action4 ”
>"And other tags.

FIG. 5 is a schematic diagram showing an outline of the in-situation knowledge. Hereinafter, description will be made with reference to this figure.

The in-situation knowledge shown in the figure is knowledge about a task, and is related to "changing a character font".

This knowledge includes information such as an activity area, a role relationship, a communication style, an interpretation construction strategy, and a partial semantic plan according to the above structure.

Also, “Partial wordin” of “wordings”
"g" means "change character font".
As “Overall wordings”, tagged natural language and everyday language expressions are stored. For example, "<task> Convert character string font </ task>".

In the present embodiment, a context-dependent dictionary 116 is provided in the semiotic base 102. The situation-dependent dictionary 116 holds the situation characteristics and the vocabulary information and the meaning information of the language under the given situation, as described above.

FIG. 6 is an explanatory diagram showing an example of vocabulary items stored in the context-dependent dictionary 116. Hereinafter, description will be made with reference to this figure.

The items shown in this figure are for "change", and include information on the situation, in addition to information such as fluctuation of notation and part of speech information.

As the activity area (field), the status value
F1_1 (question and answer) and F2_1 (word processor operation) are stored. In addition, no special situation values are recorded for role relationships (tenor) and communication style (mode).
Null is stored.

Also, SXU1 & goal, SXS & PSM, UXU1 & Previous_Ac
is also stored.

In these figures, "vocabulary" and "I want to do it" are shown as vocabulary items.

In the following, these semiotic bases 1
The process of understanding an utterance using 02 and generating a text will be described in detail.

(Understanding Process) FIG. 7 is a flowchart showing the flow of the process of the understanding process. Hereinafter, description will be made with reference to this figure.

When an utterance from the user is input (1),
Purging is performed (2), and utterance prediction confirmation is executed (3).

The following processing is performed to confirm the utterance prediction. -Matching the result of morphological analysis with the morpheme used in the corpus (4). Matching between the morphological analysis result and the context-dependent dictionary 116 (5). -Match the morpheme analysis result with the morpheme of the verbalized knowledge (6). Matching of the morphological analysis result with the concept label of the network, and matching of the result with the concept label of the knowledge corpus (7, 8)

As described above, the vocabulary of the text included in the corpus base 114, the vocabulary of the natural language and everyday language expression of the knowledge in the context included in the context base 111, the vocabulary items included in the context-dependent dictionary 116, and the vocabulary of these The situation is specified using the concept label of the item. In particular, the natural language and everyday language expressions included in the context base 111 correspond to the user's unique expressions such as expressions used in the past, so that the situation can be identified more accurately.

Thereafter, processing of resource setup (10) and shortcut search (11) are performed.
Lexical grammatical features, semantic feature identification (12), stage, move identification (13) to add meaning,
The context is understood for this and the user profile (14).

In the shortcut search (11), the context-dependent dictionary 116 can be used. As a result, lexical grammatical features and semantic features of the vocabulary under a specific situation are identified at high speed. Further, it can contribute to speeding up the identification of the stage and the move.

(Generation Process) FIG. 8 is a flowchart showing the flow of processing in the generation process. Hereinafter, description will be made with reference to this figure.

When the context is understood by the above understanding process, a situation type corresponding to the context is identified (1).

Then, various resources used in the generation process are set up (2), and the concept is activated in the concept network (3). After this, the process is split into two.

First, referring to the utterance management module, utterance type identification (4) and utterance instance identification (5)
I do.

Further, semantic features are specified from the utterance instance (6), and the context is collated with reference to the user profile in the context base 111 (7).

Next, referring to the context-dependent dictionary 116 or the like, a vocabulary to be used for the generation text is selected (8), and the connection form of the particle or auxiliary verb is specified (9).

On the other hand, a text plan is generated by referring to the in-context knowledge stored in the context base 111 and the corpus base 114 (11), and the generated plan and the user profile in the context base 111 are referred to to generate a context plan. Are collated (12).

Thereafter, a character string is generated (13), a text is generated in accordance with the specified form of the particle or auxiliary verb, and the text is distributed to the user (10).

[0086] After that, what kind of text should be understood next is predicted.
At this time, the stage base and the discourse management knowledge stored in the context base 111 can be used.

In selecting a vocabulary and specifying a particle / auxiliary verb, the use of the context-dependent dictionary 116 enables efficient and high-speed selection of a suitable vocabulary according to the current situation of the user.

In the following, a bridging process corresponding to the situation type identification will be described.

(Connection Process) FIG. 9 is a flowchart showing the flow of processing in the connection process. Hereinafter, description will be made with reference to this figure.

First, when the understanding process is completed, the process is divided into two.

First, the stage is recognized using the stage-based utterance management module in the context base 111 (1). This updates the situation type (2).

Further, utilizing the discourse management knowledge in the context base 111, the plan of the uttering user is recognized (4), and the situation type is updated (5).

In the generation process, these updated status types are used.

Now, when the status type is updated (2,
5) The utterance candidate is searched with reference to the stage base and the user profile in the context base 111 (3). Also, the listener's plan is selected with reference to the discourse management knowledge (6).

In addition, the in-situation knowledge in the corpus base 114 and the context base 111 is searched to check the role relation / communication style, and collate with the situation (8).

Thus, the utterance act of the user is identified.

In this embodiment, in the connection process, the context base 111 and the corpus base 114 are used.
Case inference using natural language and everyday language expressions stored in. This makes it possible to speed up the generation of the listener's plan and the above-described text.

[0098]

As described above, according to the present invention,
A language computer using the semiotic base 102, a language processing method, and a program for realizing these can be provided.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a schematic configuration of one embodiment of a language computer according to the present invention.

FIG. 2 is a flowchart showing a processing flow of a language processing method executed by the language computer.

FIG. 3 is a schematic view showing a structure of a semiotic base.

FIG. 4 is a schematic diagram showing an example of situation description knowledge.

FIG. 5 is a schematic diagram showing an overview of knowledge in a situation.

FIG. 6 is an explanatory diagram showing an example of vocabulary items stored in a situation-dependent dictionary.

FIG. 7 is a flowchart showing a flow of processing in an understanding process.

FIG. 8 is a flowchart showing a flow of processing in a generation process.

FIG. 9 is a flowchart showing a flow of processing in a connection process.

[Explanation of symbols]

101 language computer 102 semiotic base 103 Input reception unit 104 Situation identification unit 105 Text generator 111 Context Base 112 semantic base 113 Vocabulary Grammar Base 114 Corpus Base 115 General dictionary 116 Situation-dependent dictionary

   ────────────────────────────────────────────────── ─── Continuation of front page    (72) Inventor Hiroko Fujishiro             2-1 Hirosawa, Wako-shi, Saitama RIKEN             Inside (72) Inventor Michio Kanno             2-1 Hirosawa, Wako-shi, Saitama RIKEN             Inside F term (reference) 5B081 AA08 BB03

Claims (5)

[Claims] 1. A semiotic base, an input receiving unit,
A language computer including a situation identification unit and a text generation unit, wherein: (a) the semiotic base includes a context base, a semantic base, a lexical grammar base, a corpus base, and a general-purpose dictionary. The context base includes contextual knowledge that describes the characteristics of the context, in-context knowledge that describes the knowledge used in a given context, and a user profile that describes the issues related to the user of the language computer. Described and stored in natural language, the semantic base holds the semantic features of the language and the corresponding semantic role, the lexical grammar base holds the lexical grammatical features of the language, and the corpus base is: A language text as an example of language exchange is held together with semantic features, situational features, and lexical grammar features. (B) the input receiving unit receives an input of an utterance from the user, and (c) the situation identifying unit receives the utterance from the user who has received the input. (D) the text generator associates with the identified situation and describes the situation in a natural language and stores the context in the context base. Referring to information stored in the semiotic base including knowledge, in-situation knowledge, and a user profile, generating a text of a response to an utterance input from a user who has received the input. thing. 2. The language computer according to claim 1, wherein the text of the response generated by the text generation unit is a natural language of a situation description knowledge, a knowledge in a situation, and a user profile stored in the context base. Features that include sentences contained in the description. 3. The language computer according to claim 1, wherein said semiotic base further comprises a context-dependent dictionary, said context-dependent dictionary comprising vocabulary information of a language under a given context. And the text generating unit generates the text of the response by further referring to the vocabulary information and the semantic information of the language under the identified situation stored in the context-dependent dictionary. Characterized by that. 4. A language processing method using a semiotic base, comprising: a language processing, an input receiving step, a situation identification step, and a text generation step, wherein the semiotic base comprises: a context base;
Semantic base, lexical grammar base, corpus base,
A general-purpose dictionary, and the context base includes situation description knowledge explaining the characteristics of the situation, in-situation knowledge explaining knowledge used under a given situation, and a user profile explaining items related to the user. Described and stored in natural language, the semantic base holds the semantic features of the language and the corresponding semantic role, the lexical grammar base holds the lexical grammatical features of the language, and the corpus base is: The general-purpose dictionary holds the linguistic information and the semantic information of the language, and holds the language text as an example of the language exchange together with the semantic features, the situation features, and the vocabulary grammatical features. Receiving an utterance input from a user whose items are stored in the context base; and (b) in the situation identification step, (C) In the text generation step, a situation described in a natural language and stored in the context base in association with the identified situation. Descriptive knowledge, knowledge in context, and referring to information stored in the semiotic base including a user profile, generating a text of a response to an utterance input from a user who has received the input. how to. 5. A program for causing a computer to function as the language computer according to claim 1. Description: