JP2673831B2 - Conversational etiquette education system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention is a language and desirable conversation in accordance with social conventions, which is required in public dialogue situations such as telephone reception, customer service, interviews, and self-introduction. The present invention relates to a conversation etiquette education system using a computer device for the purpose of supporting learning of the development of.

[Prior Art] As an object close to the purpose of use of the present invention, there is an English conversation education system capable of learning English conversation while freely talking with the system. This English conversation education system has the function of pointing out the grammatical error of the user's utterance and the error about the spelling of words, and presenting the appropriate utterance wording suitable for the situation, while having free conversation in English with the user. is there. In order to realize such a function, this English conversation education system prepares knowledge such as a word dictionary, a grammar dictionary, and a concrete conversation pattern. The input of the user's conversation is a keyboard, the input sentence is displayed on the display,
The system response is output by the display and voice.

[Problems to be Solved by the Invention] In a conventional language education system, correction of an error in an input sentence is handled by a word dictionary, a grammar dictionary, and a concrete conversation pattern. The error in the conversation etiquette that is the target of the present invention is not only grammatical but also an error in the operation of honorifics, an error due to a mismatch with the expansion pattern of the conversation, etc. Can not.

[Means for Solving the Problem] As a means for recognizing and correcting an error of a type that cannot be dealt with by conventional techniques, grammatical knowledge about honorifics, knowledge for operating honorifics, conversation status information, and conversation patterns are used.

[Operation] The conversation voice input by the user is
After being recognized and understood as an input sentence by the language processing unit and the utterance understanding unit, the evaluation unit recognizes an error regarding the expression of the sentence using the honorific grammar knowledge, the honorific operation knowledge, and the conversation situation information. Furthermore, it is matched with the development pattern of the conversation,
It is determined whether the input sentence is appropriate according to the development of the conversation. The dialogue management unit generates a response of the corresponding system from the processing result and the conversation pattern. By repeating this, the system and the user have a conversation. When the conversation is completed, the comprehensive evaluation unit outputs the evaluation about the general conversation and the advice about the wording from the stored utterances of the user.

[Embodiment] FIG. 1 is a block diagram showing an embodiment of an apparatus for constructing a system of the present invention, in which 1 is a voice input / output device, 2 is a display device, and 3 is all information processing. It is a processing device for doing.

The user inputs a voice to the processor 3 using the voice input / output device 1 connected to the processor via a telephone line or directly.

The processing device 3 recognizes the voice of the user converted into an electrical signal, analyzes and understands the signal, generates a response sentence corresponding to the signal, outputs the voice to the voice input / output device, and outputs the voice to communicate with the user. Advance the conversation. Also teach conversation manners. Further, by outputting characters and image information to the display 2, the user's understanding is helped.

FIG. 2 is a block diagram showing the configuration of each processing unit in the processing device. In FIG. 2, reference numeral 4 is a recognition unit, which converts an input voice into a phoneme string by a voice recognition function. The language processing unit 5 converts the phoneme string into character information. Utterance understanding part 6
Extracts the meaning of the sentence from the character information.

The utterance understanding unit 7 receives the content of the user's utterance,
The necessary information is extracted and stored as status information, and a response as a system is generated. The honorific generation unit 8 receives the meaning content of the utterance and generates the utterance of the correct expression.

The evaluation unit 9 compares and analyzes the user's utterance converted into text information and the utterance's contents converted into sentences using a correct expression, and evaluates the user's utterance expression and contents. To do.

The comprehensive evaluation unit 10 analyzes the evaluation of the user's utterance and generates a comprehensive evaluation. The output unit 11 outputs the image information to the display, and outputs an instruction to the user of the system, a response of the system to the user's utterance, and an evaluation as voice and character information.

The situation knowledge database 14 is used by the utterance understanding unit 6, the honorific grammar knowledge database 12, the honorific operation knowledge database 13 is used by the honorific generation unit 8, and the dialogue pattern database 15 is the dialogue management unit 7 and the evaluation unit 9. Used in each. The image data is used by the dialogue management unit 7 according to the progress of the dialogue.

FIG. 3 is a flow chart showing the procedure of processing in the processing apparatus, and the reference numerals of the corresponding processing units in FIG. 2 which share each processing are shown in parentheses. The processing of the system will be described in the order of processing with reference to FIG.

There are roughly two types of processing sequences in this system, a dialogue sequence and an evaluation sequence.

First, the input speech will be recognized by the recognition unit 4 and converted into character information by the language processing unit 5 in the dialogue sequence. This character information is output from the output unit 11 as a character. Furthermore, this character information is understood by the speech understanding unit 6,
The understood content is analyzed by the interaction management unit 7 according to the interaction pattern, and the response content is generated. The response content is converted into a sentence having a correct expression by the honorific generation unit 8 and output from the output unit 11 as voice and character information. At the same time, the image data corresponding to the situation, which is obtained by the dialogue management unit 7 searching the image data, is output from the output unit 11 as an image.

Describing along the evaluation sequence, a character string having the user's utterance converted into character information in the language processing unit 5 and the utterance understanding unit 6 understands the honorific generation unit 8
The evaluation unit 9 compares and evaluates the utterance of the user having the correct expression converted in (1).

This comparison and evaluation is based on the honorific usage knowledge database that describes the human relationships required to operate honorifics for each conversation scene.
12 and a rewriteable honorific grammar knowledge database 13 in which the respectable, humbling, and polite forms of the word are described for each word. Figure 8 shows the honorific operating knowledge database 12
And a part of the honorific grammar knowledge database 13 is illustrated. Specifically, this comparison and evaluation finds error parts such as misuse and misuse of honorific words of users, misunderstanding of honorific words and humble words, saves them, and accumulates error scores for each type of error Is done by doing.

Further, the content of the utterance is evaluated by referring to the dialogue pattern. This evaluation uses a dialogue pattern 15 that describes the desirable dialogue flow for each conversation scene as a series of pairs of dialogues between the system and the user, and the content of the actual user's utterance does not deviate from this dialogue pattern 15. It is done by checking. FIG. 9 illustrates a part of the dialogue pattern 15. Specifically, when the user's utterance order and content deviate from this dialogue pattern, the deviated utterance content and the desired original utterance content are saved, and the error distance score, which is the degree of deviation of the utterance, is stored. calculate.

When the dialogue between the system and the user ends, the comprehensive evaluation unit 10
, The evaluation result of the user's utterance is comprehensively evaluated, and together with each evaluation, the output unit 11 outputs character information and voice information.
Output from The evaluation output from the comprehensive evaluation unit 10 includes an error and a correct answer in the expression of the user's utterance containing the evaluation result received from the evaluation unit 9, an indication of an undesired utterance content and a desirable correct answer example, and the evaluation unit 9 There is a comprehensive evaluation of the user's utterance derived from the score of the error type of the utterance expression received from and the error distance of the utterance content, and hints for making a correct conversation. Regarding the comprehensive evaluation and hints for correct conversation, the output contents are prepared in advance in the form of a database in the comprehensive evaluation section, and the appropriate contents are extracted using the error type score and the error distance as keys. .

FIG. 4 is a flow chart showing the operation of the system. FIG. 5 is a flow chart detailing the operation of the conversation mode shown in FIG. FIG. 6 is a flow chart showing in detail the operation of the evaluation mode shown in FIG. FIG.
6 is a flowchart illustrating in detail the operation of the learning mode shown in the figure.

The output of the system to the user will be described with reference to FIG. 4 and the associated flow charts of FIGS. When selecting the practice conversation scene, the explanation of the set situation is output to the display 2 as voice information and character information.

In the conversation mode, the device operates according to the procedure shown in the flowchart of FIG. 5, the set conversation situation is output to the display 2 as image information, and the conversation between the system and the user is output as voice and character information. Is output as.

In addition, the system determines whether the expression and content of the user's utterance is close to or far from the correct sentence expression and dialogue pattern,
The degree of the distance is represented by the output of a buzzer sound and the change in the facial expression of the human face displayed in the screen corner of the display 2.

In the evaluation mode, the device of the present invention operates in accordance with the flow shown in FIG. 6, and the evaluation regarding the content and expression of the user's conversation is output by the voice and character information, and further the comprehensive evaluation for the user's conversation is performed. Is output.

In the learning mode, which is used when the user has not learned enough conversation, the device operates according to the flow shown in FIG. 7, hints for making a correct conversation, correct conversation examples, and other possible utterances. The content and expression of
It is output as voice and text information during conversation learning with the system, which helps to learn correct conversation. Figure 10 shows a part of the conversation between the system and the user. Figure 11 shows a concrete example of the operation of conversation evaluation. Figure 12 shows an example of conversation evaluation. FIG. 13 is an example of a comprehensive evaluation of conversation, hints and advice. FIGS. 10 to 13 are examples of the evaluation of the conversation of the user of the system described in FIGS. 5 to 7.

[Advantageous Effects of the Invention] The present invention evaluates two points, that is, word usage centered on honorifics and development of conversation topics, targeting user's utterance in a specific situation set by the system, and appropriate It gives advice, and therefore, by using not only grammatical knowledge about general honorifics but also operational knowledge about honorifics, situational knowledge of the conversation world, and development patterns of conversations for each specific conversation scene, Not only the contents of conventional conversational sentences but also the problem of language usage that changes depending on the situation of language usage and topic flow can be supported by the user. The voice understands the user's utterance, and the system outputs a response sentence in a wording according to the situation of the conversation, so that the user can quickly and easily have a conversation with the system, and learn more realistically. It can be carried out.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a functional block diagram of a processing apparatus, FIG. 3 is a flow chart showing a processing procedure in the processing apparatus, and FIG. 4 shows an operation of the system. A flow chart, FIG. 5 is a flow chart explaining the operation in the conversation mode in FIG. 4, FIG. 6 is a flow chart explaining the operation in the evaluation mode in FIG. 4, and FIG. 7 is a flow chart explaining the operation in the learning mode in FIG. It is the flow chart which it did. Further, FIG. 8 is an example of contents of the honorific operating knowledge database and honorific grammar knowledge database, FIG. 9 is an example of dialogue pattern, FIG. 10 is an example of a part of conversation between the system and the user, and FIG. 11 is A concrete operation example of conversation evaluation, FIG. 12 is an example of conversation evaluation, and FIG. 13 is an example of comprehensive evaluation of conversation, hints, and advice. 1 ... Voice input / output device, 2 ... Display device 3 ... Processing device, 4 ... Recognition unit 5 ... Language processing unit, 6 ... Utterance understanding unit, 7 ... Dialog management unit, 8 ... Honorifics generation unit 9 …… Evaluation section, 10 …… Comprehensive evaluation section 11 …… Output section, 12 …… Honorific grammar knowledge 13 …… Honorific operation knowledge, 14 …… Situational knowledge 15 …… Dialogue pattern, 16 …… Image data

Claims (1)

(57) [Claims] 1. A speech recognition unit for recognizing an input user's voice as a phoneme and a character string, a language processing unit for generating an input sentence from the recognized character string, and an understanding of the input sentence in the context of a conversation. Conversation comprehension section, grammatical knowledge about honorifics, information about conversation situations, knowledge for operating honorifics according to conversation situations, and large-capacity storage device as a database that stores information about conversation patterns for conversation progress, read from database The evaluator evaluates the expressions and contents of the conversational sentences input using information on the honorific grammar, the knowledge of honorific operation, conversation situation information, and conversation patterns. Corresponding to the input sentence by referring to the conversation situations and conversation patterns The dialogue management unit that generates the response, the honorific generation unit that gives the correct expression according to the situation of the generated response sentence, and analyze each evaluation of the user's conversation sentence Comprehensive evaluation unit that determines the educational policy for learning manners and outputs learning advice to the user, and output unit that outputs the response voice corresponding to the input voice and character image information to support the user's learning A conversation etiquette education system equipped with each device.