JP2019185230A - Conversation processing device and conversation processing system and conversation processing method and program - Google Patents

Abstract

To respect a user as well as continuing conversation getting close to feeling of the user and lead spontaneous resolution of the use's trouble and problem while resolving the user's stress/loneliness/anxiety by taking out the user's inner world.SOLUTION: A conversation processing device 1 comprises an output response sentence creation unit 60 that makes at least a part of natural language texts included in user's utterance an input sentence and creates an output response sentence based on one or more of response sentences of either an input sentence with a response sentence for confirmation facilitation where a response sentence for confirmation facilitation that facilitates confirmation of a contents of user's utterance is connected to an ending of the input sentence, a response sentence for sympathy showing sympathy on the user's utterance, and a response sentence for conversation contents embodiment that embodies a contents of a conversation with the user.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a conversation processing apparatus, a conversation processing system, a conversation processing method, and a program.

  According to a survey conducted by the Ministry of Health, Labor and Welfare, the number of patients with mood [emotion] disorders has exceeded 1 million in 2008 and has reached about 1.12 million in 2014. Is increasing. The user may not be able to verbalize the essence of his / her problems and issues, and in that case, it is difficult to solve the issues and problems with his / her own power.

  As described above, in order to solve the problems and problems that the user has, it is conceivable that the user can receive support by talking with a counselor (consultant) or receiving counseling. However, there are still few counselors who are familiar with the problems and issues of users, and the number of counselors that can be handled by the counselor is about 20 to 30 per month.

Therefore, there is a need for a software agent that can act as a user support by a counselor. For example, ELIZA disclosed in Non-Patent Document 1 creates a response sentence that returns a so-called parrot that rephrases the text content input by the user with the same text content.
In addition, in order to continue the conversation, it also makes a response that guides it to a new topic.
However, since ELIZA responds to the same content as the user's input content and expands the topic to continue the conversation, it focuses on the problems and issues faced by the user, Solving the problem is considered difficult.

  In order to solve the above-mentioned deep mental problems and problems of users, an introspection support apparatus described in Patent Document 1 has been proposed. The introspection support device described in Patent Literature 1 associates occupations with tasks corresponding to occupations, and stores a storage device that stores attribute data that associates occupations and related keywords of the tasks, and identifies user issues The apparatus includes problem identification means and problem solution support means for assisting the user in solving the identified user problem. The introspection support device described in Patent Document 1 uses the user's spoken or emotional keywords, maintains the context, and repeats the response by interlacing inquiries to inquire about the details of the task. Identify worries and issues, encourage users to organize by introspection, and give improvement awareness.

JP 2014-229180 A

Yoshinobu Kano; "Can we communicate with computers: the present state of dialogue systems", Information Management, Vol. 59, No. 10, pp. 658-665, (2016)

  As described above, the software agent described in Non-Patent Document 1 has a problem that it is difficult to materialize the troubles and problems that the user has and to solve deep mental problems and problems. In addition, in the introspection support apparatus described in Patent Document 1, an inquiry is made to the user to identify a job and a problem corresponding to the job, but an enormous amount of knowledge information is required for an appropriate query. The construction of the device is expensive. In addition, depending on the user's problems and issues, it is often difficult to make appropriate inquiries, and the user feels the story has changed and the user feels stress and anxiety when the inquiries are inappropriate or the inquiries themselves. There was a problem. In addition, in the introspection support device described in Patent Document 1, a response sentence is created by focusing on problem identification, while the user thinks that he / she is not respected or does not understand his / her position There was also a problem that sometimes I felt loneliness. In addition, such a sense of distance between the user and the introspection support device is affected, and it is difficult for the conversation to last for a long time, and the user's deep troubles are not voluntarily resolved. There was a problem.

  The present invention has been made in order to solve the above-mentioned problems, and respects the user and keeps talking while keeping close to the user's feelings. Provided are a conversation processing device, a conversation processing system, a conversation processing method, and a program that can lead to a user's troubles and problems to be resolved spontaneously while eliminating anxiety.

    In order to solve the above-described problem, an aspect of the present invention provides a confirmation promoting function that uses at least a part of a natural language element included in a user's speech as an input sentence and promotes confirmation of the content of the comment at the end of the input sentence Any one of an input sentence with a confirmation prompting response sentence connected with a response sentence, a response sentence for empathy indicating empathy for the utterance, and a response sentence for recognizing the conversation content that embodies the content of the conversation with the user A conversation processing apparatus including an output response sentence creating unit that creates an output response sentence based on a response sentence including one or more.

  Further, according to one aspect of the present invention, the output response sentence creation unit creates the output response sentence by omitting any one or more of second-person expressions, modifiers, and predicates that indicate the user itself, The output response sentence includes at least an input sentence with a confirmation prompting response sentence and a response sentence for specifying the conversation content, and the output response sentence creating unit includes the input sentence with a confirmation prompting response sentence, the conversation A conversation processing apparatus that creates the output response sentence including a response sentence for content instantiation.

  In addition, according to one aspect of the present invention, the presence or absence of the extraction of the natural language element from the utterance is determined based on the input of the user, and the utterance is re-sent to the user when it is determined that the defect exists. Is a conversation processing device that further includes a natural language element extraction failure determination unit.

  Further, according to one aspect of the present invention, the above-described conversation processing device, an input sentence storage unit that extracts the natural language element included in the speech and stores at least a part of the natural language element as the input sentence; A log storage unit for recording past conversation contents of a user including the input sentence, the response sentence for empathy, the response sentence for confirmation confirmation connected to the input sentence and the input sentence, and the content for conversation contents A response sentence storage unit in which a response sentence is stored, and the output response sentence creation unit reads the input sentence from the input sentence storage unit and the log storage unit, and the empathy from the response sentence storage unit The conversation processing system is characterized in that the response sentence for response, the input sentence, the confirmation prompting response sentence connected to the input sentence, and the response sentence for realizing the conversation content are read out and the output response sentence is created.

  One embodiment of the present invention is a conversation processing system further comprising: an input device to which the above-mentioned user's speech is input; and an output device that outputs the output response sentence as audio information and visual information.

    Another aspect of the present invention is a conversation processing system in which the input device acquires environment information on an input side including a user, and the output response sentence creation unit creates the output response sentence based on the environment information. .

  Further, according to one aspect of the present invention, the input sentence storage unit and the log storage unit store the user's utterance at the time of past conversation as an already input sentence, and the output response sentence creation unit includes the input sentence as the input sentence. It is a conversation processing system that creates the output response sentence including a past response sentence indicating the past, the already input sentence, and the confirmation prompting response sentence when it is determined to match the already input sentence.

  One embodiment of the present invention is a conversation processing system further including at least one of a head, a torso, a hand, and a leg that moves according to the output response sentence.

  One embodiment of the present invention is a conversation processing system in which any one or more of the head, the body, the hand, and the leg are virtually represented.

  Further, according to one aspect of the present invention, an input sentence obtaining step for extracting a natural language element included in a user's utterance and obtaining an input sentence, and confirmation promotion for promoting confirmation of the content of the utterance at the end of the input sentence Any of an input sentence with a confirmation prompting response sentence to which a response sentence for confirmation is connected, a response sentence for empathy indicating empathy for the utterance, and a response sentence for embodying a conversation content that embodies the contents of the conversation with the user And an output response sentence creating step of creating an output response sentence using a response sentence including one or more.

  Further, according to one aspect of the present invention, an input sentence obtaining step for extracting a natural language element included in a user's utterance and obtaining an input sentence, and confirmation promotion for promoting confirmation of the content of the utterance at the end of the input sentence Any of an input sentence with a confirmation prompting response sentence to which a response sentence for confirmation is connected, a response sentence for empathy indicating empathy for the utterance, and a response sentence for embodying a conversation content that embodies the contents of the conversation with the user An output response sentence creating step for creating an output response sentence using a response sentence including one or more.

  According to the present invention, the user continues to have a conversation while respecting the user's feelings and pulling out the inner surface of the user, thereby relieving the user's stress, loneliness and anxiety, while resolving the user's troubles and problems. It can lead to a spontaneous solution.

It is a block diagram which shows an example of the conversation processing system by the 1st Embodiment of this invention. It is a flowchart which shows the procedure of the conversation process by a conversation processing apparatus and a conversation processing system. It is the example of a conversation by a conversation processing apparatus and a conversation processing system. It is a block diagram which shows an example of the conversation processing system by the 2nd Embodiment of this invention. It is a schematic diagram of a part of the robot apparatus of the conversation processing system.

  Hereinafter, a conversation processing device, a conversation processing system, a conversation processing method, and a program according to a first embodiment of the present invention will be described with reference to the drawings.

[First Embodiment]
(Conversation processing device, conversation processing system)
FIG. 1 is a functional block diagram illustrating an example of a conversation processing system according to the first embodiment. As shown in FIG. 1, the conversation processing system 100 includes at least a conversation processing device 1. The conversation processing apparatus 1 includes at least an output response sentence creation unit 60, and includes a speech recognition unit 12, a determination unit (natural language element extraction failure determination unit) 80, a response sentence selection unit 94, a learning unit 92, and a speech generation unit 13. Further prepare. The conversation processing system 100 includes an input device 2, a storage device 3, and an output device 4 in addition to the conversation processing device 1. The storage device 3 includes an input sentence storage unit 30, a response sentence storage unit 40, and a log storage unit 91.

  The input device 2 is a device that can input a user's speech, and examples thereof include a microphone, a keyboard, and a digital pen. The input form of the user's speech to the input device 2 is not particularly limited as long as a natural language element included in the speech can be extracted, and widely includes a user's voice, characters representing the content of the user's speech, and the like. In the following description, it is assumed that the user's voice is input to the input device 2. However, when the user's utterance is input to the conversation processing device 1 in an input form other than voice, the voice recognition unit 12 and the voice generation unit 13 are replaced with a recognition unit and a generation unit suitable for an input / output mode other than voice. .

The input device 2 may have a function of acquiring environment information on the input side including the user in addition to the above-described functions. The environment information on the input side including the user includes, for example, information in the non-verbal communication such as the user's facial expression, face color, voice color, number of blinks, mouth movement, gesture gesture, gaze, etc. during conversation, body temperature, pulse wave, It is biological information such as brain wave fluctuations. The means by which the input device 2 acquires environmental information is not particularly limited, and examples thereof include a line-of-sight tracking device, an infrared sensor, a thermometer, a depth sensor, a seating sensor, a pulse wave sensor, and an electroencephalogram sensor.
Based on body information such as line of sight, facial expression, body movement, and utterance change acquired by the input device 2 and biological signals such as fever, pulse, and brain waves, the output response sentence creation unit 60 estimates the user's emotion and its change. The method for estimating the emotion is not particularly limited, and a known method can be used.

The input sentence storage unit 30 extracts natural language elements included in the user's statement and stores at least a part of the extracted natural language elements as an input sentence. The input sentence storage unit 30 receives a natural language element included in the user's voice (utterance) from the input device 2 via the voice recognition unit 12, for example. Specifically, the input sentence storage unit 30 is a computer memory or the like. The input sentence storage unit 30 may be a volatile memory that temporarily stores data, a non-volatile memory that stores data for a long time, a hard disk, or the like, or may be a storage medium on a network server.
Further, the input sentence storage unit 30 may also function as a log storage unit 91 described later. Unless otherwise specified, the input sentence storage unit 30 will be described as also functioning as the log storage unit 91. .

  The voice recognition unit 12 performs voice recognition on an arbitrary voice recognition server for the user's voice input from the input device 2 and converts the user's voice into natural language text (natural language elements). The converted natural language text is subjected to predetermined checks and corrections.

The speech recognition unit 12 performs, for example, morphological analysis, syntactic analysis, and syntax analysis of an input sentence of the input user's speech (utterance) in order to perform natural language processing such as rephrasing confirmation of the main point of the user's speech. When the input sentence is Japanese, the tools for performing morphological analysis and the like are not particularly limited. For example, MeCab (http://taku910.github.io/mecab/), KAKASHI (http://kakasi.namazu.org/ ) And other standard tools can be used. An example of morphological analysis for the user's voice (utterance) will be shown below.
<Example 1>
User voice (speech): “I'm worried about job hunting.”
→ (Job hunting, Noun, Shushoku Katsudo) (de, particle, de) (worse, verb, nayan) (de, particle, de) (i, verb, i) (mas, auxiliary verb, mass)
<Example 2>
User voice (speech): “I want to work for SE”
→ (self, noun, jibn) (ha, particle, ha) (SE, noun, sue) (no, particle, no) (job, noun, shigoto) (do, verb, shi) (tai, auxiliary verb, thai)

In the input sentence storage unit 30, the natural language text extracted from the user's voice by voice recognition using the voice recognition unit 12 is an input sentence for creating an output response sentence (hereinafter referred to as an input sentence for creating a response sentence). Remember as).
Further, tags indicating the purpose, desire, means, reason, consequent emotion, etc. are added to the natural language text extracted from the user's voice and stored in the input sentence storage unit 30.
When there is environmental information acquired in synchronization with the user's voice, a tag such as emotions of emotions is further added and stored in the input sentence storage unit 30.

  The determination unit 80 determines whether or not the natural language text extracted from the user's voice is defective. If there is a defect, the predetermined correction described above is performed. Judgment on extraction of natural language text is, for example, a recursive neural network (RNN) or a Japanese sentence predictor using N-gram (for example, “Flick: Japanese Input Method Editor using N-gram and Recurrent Neural Network Language Model based Predictive Text Input”) , Yukino Ikegami, Yoshitaka Sakurai, Ernesto Damiani, Rainer Knauf, Setsuo Tsuruta: SITIS2017, Jaipur, India, December 4-7, 2017, etc.).

  The determination unit 80 may provide a determination opportunity regarding the presence or absence of extraction failure of the natural language text extracted from the user's voice, and may make a determination based on the input from the user. In this case, the determination unit 80 causes the output device 4 described later to display natural language text extracted from the user's voice. The user determines whether or not the displayed natural language text is defective, and inputs the determination result via the input device 2. When the determination unit 80 determines that there is a defect by the user, the determination unit 80 performs the predetermined correction described above.

  When it is determined that there is a defect in the extraction of the natural language text, the output response sentence creating unit 60 generates a natural language text such as “Please speak again” as an output response sentence. When outputting the output response sentence as voice information, the voice generation unit 13 converts the output response sentence into voice information, and outputs the voice response information to the user as voice information or visual information through the output device 4. Can be requested. However, from the viewpoint of response time, load, and software ecology, the speech recognition unit 12 recognizes and interprets speech, that is, misunderstanding, etc. before processing in the output response sentence creation unit 60 from the viewpoint of response time, load, and software ecology. It is preferable that the output response sentence creation unit 60 outputs the childishness or the superficial input failure having a low relevance to mental, cognitive, philosophical awareness, or the like as visual information.

  In addition, in order to process an input failure that can only be determined by the user who has a mental problem or a person in contact with the user, the determination unit 80 displays the input result as visual information on a part of a display such as a confirmation window. For example, you may make it wait for a fixed time for 2 to 10 seconds. When the user confirms an input failure, the user inputs again through the input device 2.

  When the user performs re-input, the input device 2 may be configured to allow the user to re-input by pressing a switch or button provided in the input device 2.

  The speech recognition unit 12 may use a conversion method or speech recognition method different from the previous extraction for the re-input user's speech. Examples of different conversion methods include alternative conversion table items, conversion dictionaries, and examples of different speech recognition methods include alternative speech recognizers and multiplexed speech discriminators. Different conversion methods may be changed depending on the input failure or the number of re-inputs.

In addition, the output response sentence creation unit 60 is in the waiting time in order to prevent misrecognition due to the limits of the computing power and processing power of the conversation processing system 100 itself and the interruption of the conversation due to the waiting time for the confirmation. Alternatively, before and after that, audio output and visual information of the second embodiment described later may be output, and the above-described audio and visual information may be output as a re-input promotion output.
Re-input / transmission output refers to output that displays voice or operation such as “Humhum”, text on a part of the display such as a confirmation window, or operation such as whispering. is there.

  The determination result regarding the presence or absence of the natural language text extraction failure in the determination unit 80 and the correction result by dictionary use / predictive automatic correction / conversation correction are stored in the input sentence storage unit 30. In addition, the determination result and the like are displayed on the output device 4.

  In addition, the determination unit 80 can perform automatic confirmation of failure by a return value from the speech recognition unit 12 and retry speech recognition in the speech recognition unit 12 at the time of failure. It is possible to confirm the conversation by displaying the determination result (that is, re-input) and retry the speech recognition in the speech recognition unit 12 at the time of failure. The correction result is stored in the input sentence storage unit 30 and reflected in a dictionary corresponding to each user.

In the response sentence storage unit 40, a response sentence for empathy indicating empathy for the user's remarks, a confirmation prompting response sentence for promoting confirmation of the remark contents that are the contents of the remarks of the user immediately before, and a certain period with the user A conversation content instantiation response sentence that embodies the conversation content that is the content of the user's remarks multiple times is stored. As the response sentence for empathy, for example, “Fun”, “Fumufum”, “Yunun” and the like can be mentioned. The response message for empathy contributes to the conversation processing device 1 and the conversation processing system 100 respecting and respecting the user and talking to the user when talking with the user. Examples of the confirmation prompting response include “what is it” and “what is it”. The confirmation prompting response sentence has an effect of making the user naturally aware of the contents that the user has spoken. Examples of the response text for specifying the conversation content include “So”, “Please tell me a little more”, “Please tell me more specifically”, “More details”, and the like. The response message for instantiating the conversation content has an effect of drawing out the inner surface of the user without departing from the gist of the user's speech or the content of the speech.
Moreover, the response sentence for instantiating the conversation content also has a function as a response sentence for problem focusing that has an effect of narrowing the focus of the troubles and issues that the user has and making it easier for the user to notice the problem.

In addition, the response sentence for empathy, the response sentence for confirmation promotion, and the response sentence for specifying the conversation content may be concise so as not to make the user feel uncomfortable in any context and not to distract the topic. preferable.
In addition, it is preferable to prepare a plurality of response sentences for empathy, response sentences for promoting confirmation, and response sentences for specifying conversation contents even if they have similar meanings so as not to bore the user.
In addition, the response sentence for empathy, the response sentence for confirmation promotion, and the response sentence for specifying the conversation content may be a sentence, or may be one or more words.
Moreover, the response sentence for empathy, the response sentence for confirmation promotion, and the response sentence for actualizing the conversation content may be selected from a plurality of types based on a predetermined rule.

  The output response sentence creation unit 60 reads the input sentence from the input sentence storage unit 30 and reads the response sentence for empathy, the response sentence for confirmation promotion, and the response sentence for specifying the conversation content from the response sentence storage unit 40. The output response sentence creation unit 60 includes an input sentence with a confirmation prompting response sentence in which a confirmation prompting response sentence is connected to the end of the read input sentence, a response sentence for empathy, and a response sentence for specifying the conversation content. An output response sentence is created using any one or more response sentences. The output response sentence creation unit 60 creates an output response sentence including an input sentence with a confirmation prompting response sentence as a basic form. The output response sentence creation unit 60 creates an output response sentence including an input sentence with a confirmation prompting response sentence and a response sentence for recognizing a conversation content as an expanded form. Furthermore, the output response sentence creation unit 60 creates an output response sentence to which a response sentence for empathy is added as another development. Since the user feels familiarity during the conversation, the response message for empathy is preferably added to the beginning of the sentence.

As described above, the output response sentence creating unit 60 extracts the natural language text from the user's remarks. From the extracted natural language text, the user himself / herself such as “I am”, “I am”, “I am”, etc. Represent the main purpose of the statement by omitting the second person expression phrase (hereinafter sometimes simply referred to as the second person expression phrase) or the adjective, adverb modifier, predicate including verb, etc. It seems to be Japanese and may be simplified.
For example, when an output response sentence is created by omitting a second person expression phrase, in the above-described morphological analysis example <example 2>, first person expression phrases that can be rephrased as a second person expression phrase are (self) and (ha). It is.

Specifically, for a morphological analysis result of a user's voice (speech) using the speech recognition unit 12, any one of a second person expression phrase, an adjective, a modifier such as an adverb, a predicate including a verb, or the like Logging is performed to create a response sentence for confirming the intention of the user's voice (speech) immediately before excluding a plurality and for narrowing-down conversation continuation. Such logging can be performed, for example, by selecting and executing a conditional firing rule exemplified in <Example 3> using regular matching.
<Example 3>
User voice (speech): “I want to work for SE but I am worried that I can get a job because I am not good at the program.”
Output response: “I want to work for SE but I'm not good at the program, so I'm worried about getting a job. Please tell me a little more.”
Condition firing rules:
・ Conditional statement ...
[ur'I am ((?: SE) | (?: PM) | (?: Programmer) | (?: System developer) @ 1) (. * @ 2) (?: Job) | (?: Join) | (?: Graduation) | (?: Pass) @ 4) (. * @ 5) ((?: Anxiety) | (?: Worry) | (?: Attention ) @ 6) (. * @ 7) '
(Note that the above-mentioned @n is an explanatory description that does not exist in the coding sentence, u is Unicode, and r is a regular expression.)
・ Log sentence ...
[u "Yeah% 1% 2 I want to get% 4 [Result]% 5% 6 because [Purpose] is% 3 [Disability]. Please tell me a little more.",
u "Fumumum% 1% 2 [Purpose] is% 3 [Disability] so% 4 [Result]% 5% 6% 7. Please tell us more about% 4% 5% 6.,
u "[Target] is% 3 [Disability] for% 1% 2, so% 4 [Result]% 5% 6% 7. Please tell us specifically about% 4% 5% 6 because% 3."]
In addition, as in <Example 3> above, using a confirmation prompting response sentence such as “I like” to facilitate the user's own awareness, respecting the user who is speaking and Emphasize.
By extracting a word indicating the main point from the input sentence by such an abbreviation process and a response sentence adding process, and constructing an output response sentence indicating the main point, the user can naturally talk without stress, and the input sentence Impress or enhance confirmation to enhance empathy, continuation and awareness.

As mentioned above, second-person expressions, adjectives, modifiers such as adverbs, predicates containing verbs, etc., and output response sentences with confirmation-promoting response sentences are not flowchart-like procedural execution, for example, by regular matching. It is created by firing and executing the condition firing rule. The output response sentence creation unit 60 creates an output response sentence by constructing a so-called rule-based engine. An example of the output response sentence described above is shown below.
<Example 4>
User voice (speech): “I want to be SE but I ’m worried because I ca n’t program.”
Output response: “I want to be SE but I'm uneasy because I can't program.” / “I want to be SE but I'm uneasy because I can't program.”
<Example 5>
User voice (speech): “I liked him.”
Output response: “You like him.”
<Example 6>
User voice (speech): “I want to be SE, but I am worried because my senior told me that“ you can't program because it ’s tough ”.
Output response: “I want to be SE, but I am worried because my senior told me“ I ca n’t program ”, so I ’m worried.” / “I want to be SE, but my senior says“ I ca n’t program because I ca n’t program ”. I am worried because I was
<Example 7>
User voice (speech): “I want to be SE but I ’m worried because I ca n’t program.”
Output response: “Fumufu, I want to be SE, but I ’m worried because I ca n’t program. Please tell me a little more.”

The output response sentence creating unit 60 can create an output response sentence by omitting redundant modifiers in addition to not using the second person expression phrase. An example of such an output response sentence is shown below.
<Example 8>
User voice (speech): “I want to join Company A, but I ’m worried because I ca n’t write the program very well.”
Output response: “I want to join Company A, but I ’m worried because I ca n’t write a program.”
In <Example 8>, in the input sentence storage unit 30, the user's voice (speech) is “I want to get into Company A [Purpose] does not work well (write) [Reason] Anxiety [Consequence]]. ”And is stored as an input sentence stating that“ I want to join Company A, but I am worried because I cannot write a program ”.

In the input sentence storage unit 30 and the output response sentence creation unit 60, there is a log storage unit 91 that can store the contents and number of output response sentences, evaluation data including the number of responses, and response states including the types of output response sentences as a log. It is connected. The log storage unit 91 includes a past log record (that is, a natural language text included in a past voice) corresponding to each user and a dictionary log that records emotion words. In addition, the output response sentence creation unit 60 is provided with a learning unit 92 that learns an effective response method based on the above-described evaluation data and the response method of the output response sentence.
The log storage unit 91 may also function as the input sentence storage unit 30.

The output response sentence creation unit 60 refers to a dictionary log or the like in the log storage unit 91 and extracts changes in the content of the user's voice (speech) over time from past conversation records in the input sentence storage unit 30. The output response sentence creation unit 60 creates an output response sentence that presents a change if there is a change including the appearance of the first or different emotion word within a predetermined number of exchanges (N1 times) of past conversations. In addition, the output response sentence is added to the logging to the log storage unit 91. That is, the input sentence storage unit 30 stores the natural language text included in the user's voice (speech) in the past conversation as an already input sentence in the log storage unit 91, and the output response sentence creation unit 60 stores the user's voice ( When it is determined that the input sentence based on (speech) matches any of the already input sentences, an output response sentence including a past response sentence indicating the past, an already input sentence, and a confirmation prompting response sentence is created. An example of such an output response sentence is shown below.
<Example 9>
User voice (speech): “I want to be SE but I ’m worried because I ca n’t program.”
Output response statement: “You said before,“ I want to become an SE but I am worried because my senior told me that it was difficult because I couldn't program. ”
If the purpose of the previous statement is stated within N3 conversations, create and output an output response sentence that presents the current (ie, immediately preceding) statement and the response for continuing the conversation (for topic specificization) The response sentence is added to the logging to the log storage unit 91.
In addition, when the output response sentence creating unit 60 refers to the log storage unit 91 and determines that the input sentence matches, for example, an already input sentence representing the purpose, the user issues a sentence before the already entered sentence representing the purpose. It is also possible to create an output response sentence including related input sentences.

If the desire (that is, the purpose and the task) is included even if the gist of the user's voice (speech) does not change, the output response sentence creation unit 60 performs the conversation after exchanging conversations N2 times or more. An output response sentence that presents the relationship with the emotion is created and added to logging to the log storage unit 91. When the content of the user's voice (speech) includes a failure or achievement means in achieving the desire, the output response sentence creating unit 60 creates an output response sentence that also presents the relationship between the achievement of the desire and the obstacle or the achievement means. In addition, the output response sentence is added to the logging to the log storage unit 91.
An example of such an output response sentence is shown below.
<Example 10>
User voice (speech): “I want to join Company A, but I ’m worried because I ca n’t develop the system.”
Output response: “Hum Fumu, I want to join Company A, but I ’m worried because I ca n’t develop the system. Please tell me a little more.”
User voice (speaking): “It's slow to make a program”
Output response: “Yeah, it ’s slow to make a program. More specifically.”
User voice (speaking): “Submissions are usually tedious during program exercises”
Output response: “Submission is not possible in the program exercise and system development is related.

  In addition, in <Example 10>, the output response sentence creation unit 60 may perform a process of replacing some words with “it”, “that”, “this”, etc. in order to emphasize the main point briefly. For example, the process of substituting “I can't do system development because I can't do system development” is replaced with “I can't do system development.

  Between the response text storage unit 40 and the output response text creation unit 60, a response text selection unit 94 may be provided. The response sentence selection unit 94 uses a plurality of types of empathy response sentences and conversation contents stored in the response sentence storage unit 40 as response messages for empathy and response sentences for embodying conversation contents used for output response sentences. The response sentence for instantiation is selected based on a predetermined rule, and each selected response sentence is output to the output response sentence creating unit 60. The predetermined rule may be a fuzzy rule, for example. Depending on the number of responses in the conversation of various output response sentences based on the above-mentioned rules, the type of response sentence for empathy in the output response sentence, the presence / absence or type of the response sentence for specifying the conversation content, Make changes to the agenda, purpose, etc. The order of output response sentences can be controlled by a learning machine or learning dictionary such as probability, fuzzy rule, neural network, deep learning, genetic algorithm, and the like. Moreover, you may select an output response sentence using the information of the tag added to the natural language text. In this way, by appropriately selecting the response sentence for empathy and the response sentence for instantiating the conversation content used for the output response sentence, the response sentence for empathy and the response sentence for instantiating the conversation content are less likely to be used repeatedly. Since the user is less likely to be bored and feel less stressed, conversation is promoted.

The voice generation unit 13 converts the output response sentence output from the output response sentence creation unit 60 into voice information, and outputs the voice information to the output device 4.
The voice generation unit 13 may convert the output response sentence into voice information in which the sound quality difference or the inflection difference is changed according to the content or the number of responses. Thereby, the action to a user's emotion can be strengthened.

The output device 4 may output visual information such as the same characters as the audio information, or may output the audio information and the visual information at the same time.
The output device 4 is a device for outputting audio information or visual information so that the user can perceive it.
The output device 4 outputs the audio information output from the audio generation unit 13 as audio. In this case, the output device 4 is constituted by a speaker, for example. Moreover, when outputting the output response sentence output from the output response sentence preparation part 60 as visual information, such as a character, the output device 4 is comprised with a liquid crystal monitor, for example.

(Conversation processing method, program)
The conversation processing method according to the first embodiment is executed by the conversation processing apparatus 1 and the conversation processing system 100 described above. The conversation processing apparatus 1 and the conversation processing system 100 described above are configured by a computer in which the conversation processing program of the first embodiment is installed, and function as a user supporter (agent).

  Next, a procedure for processing a conversation by the conversation processing device 1 and the conversation processing system 100 will be described. FIG. 2 is a flowchart showing a procedure of conversation processing by the conversation processing device 1 and the conversation processing system 100. FIG. 3 is a diagram for explaining a conversation example in the conversation processing method according to the first embodiment.

  First, when the user activates the conversation processing apparatus 1, the conversation processing system 100, the conversation processing program, and the like, the process proceeds to step S11. In step S <b> 11, the user speaks about his / her problems and issues toward the input device 2. In the conversation example shown in FIG. 3, the user says the sentence PH51 “I want to get a doctorate in 3 years, but I can't write a paper, so I ’m worried”. However, if the user does not speak within a predetermined time during the first conversation, the output response sentence creation unit 60 generates an output response sentence such as “Please tell me something you are interested in” and generate a voice. The unit 13 converts the voice information into voice information, and outputs the voice to the user through the output device 4.

  Next, the user's speech is read by the input device 2 and input to the voice recognition unit 12 and voice recognition is performed, and natural language text is extracted. It is determined whether the natural language text transmitted from the voice recognition unit 12 is received by the determination unit 80 through the output response sentence creation unit 60 (step S12). As a result, when it is determined that the natural language text has not been received or that another error has occurred (step S12: NO), the determination unit 80 again performs the user unless there is a forced termination instruction or the like from the user. Ask for remarks.

  If it is determined in step S12 that the natural language text has been received (step S12: YES), the determination unit 80 adds an appropriate log or the like to the natural language text extracted from the user's remarks, and the natural language after the log is added. The language text is output as an input sentence from the voice recognition unit 12 to the input sentence storage unit 30 through the output response sentence creation unit 60 (step S14).

  If it is determined in step S12 that the natural language text has been received (step S12: YES), the natural language text received from the speech recognition unit 12 is “clean”, “motivated”, “I feel like I can do it”. It is determined whether or not an end fixed sentence indicating that the user has solved the problem or problem is not included (step S13).

  If it is determined that the natural language text received from the speech recognition unit 12 does not include the end fixed sentence (step S13: YES), the input sentence is temporarily stored in the input sentence storage unit 30 in step S14, and then the step As S14, it is determined whether or not the already-input sentence is included in the natural language text received from the speech recognition unit 12 (step S15). In step S15, when it is determined that the natural language text does not include the already input word (step S15: YES), the input sentence storage unit 30 outputs the input sentence to the output response sentence creation unit 60. The response sentence storage unit 40 that stores a plurality of response sentences for empathy, a response sentence for confirmation promotion, and a response sentence for specifying the conversation content in advance outputs various response sentences to the output response sentence creation unit 60. The output response sentence creation unit 60 creates an input sentence with a confirmation prompting response sentence, and includes at least one of the input sentence with the confirmation prompting response sentence, the response sentence for empathy, and the response sentence for specifying the conversation content. An output response sentence is created using the sentence (step S16).

  In the conversation example shown in FIG. 3, the output response sentence OH51 “Fun, hmm. I want to get a doctoral degree in three years, but I ca n’t write a paper, so I ’m worried.” Yes. The output response sentence OH51 includes a response sentence for empathy, an input sentence with a confirmation prompting response sentence, and a response sentence for specifying conversation contents in this order.

  In step S14, if the natural language text received from the speech recognition unit 12 includes a previously input sentence that matches in the tag or a related input sentence such as a purpose (aspiration) and means (failure / reason). When the determination is made (step S15: NO), the output response sentence creation unit 60 creates an output response sentence including a past response sentence such as “before” indicating the past, an already input sentence, and a confirmation prompting response sentence (step). S17).

  In any one of steps S16 and S17, the output response sentence creation unit 60 transmits the created output response sentence to the voice generation unit 13, and the voice generation unit 13 converts the received output response sentence into voice information. The converted voice is uttered from the output device 4 to the user (step S18). Thereafter, since the user's next statement is required, the process returns to step S11. Thereafter, steps S11 to S18 are repeated as appropriate.

  When the user has solved his / her problems and problems while repeating Step S11 to Step S18 (Step S13: NO), the output response sentence generator 60 does not proceed to Step S14, but ends the conversation. An end response sentence such as “It was good” or “Thank you for your work” is output to the voice creating unit 60 (step S19), and the conversation process is terminated.

  As described above, the conversation processing device 1 and the conversation processing system 100 according to the first embodiment are any one of the input sentence with the confirmation prompting response sentence, the response sentence for empathy, and the response sentence for specifying the conversation content. An output response sentence is created using the above response sentence. Therefore, while respecting the user and staying close to the user's feelings, continuing the conversation without diverting the story and eliciting the user's inner surface, the user's troubles, This can lead to the spontaneous resolution of issues. In the conversation example shown in FIG. 3, after the first sentence (speech) PH51 of the user, the sentence (PH52, PH53, PH54, PH55, PH56, PH57) issued by the user from the agent (that is, the conversation processing device 1, the conversation processing system 100). On the other hand, the output response sentences OH51, OH52, OH53, OH54, OH55, OH56, and OH57 can be responded without including a response sentence that is different from the main idea of those sentences, and the inner surface of the user can be enhanced.

  Further, in the conversation processing device 1 and the conversation processing system 100 according to the first embodiment, the output response sentence creation unit 60 omits at least one of the second person expression phrases, modifiers, and predicates indicating the user himself / herself. Since the output response sentence is created, the response sentence can be simplified and a more natural conversation can be exchanged.

  Further, in the conversation processing device 1 and the conversation processing system 100 according to the first embodiment, the output response sentence creating unit 60 creates an output response sentence including an input sentence with a confirmation prompting response sentence and a response sentence for specifying the conversation content. Therefore, without exchanging contents other than the user's remarks, it is possible to emphasize the content of the remarks and to further draw out the inner surface of the user.

  Further, in the conversation processing device 1 and the conversation processing system 100 according to the first embodiment, a response message for empathy used for an output response message and a response message for specifying a conversation content are a plurality of types of response messages for empathy, and The response content selecting unit 94 further selects a response content for specifying the conversation content based on a predetermined rule. According to such a configuration, the response sentence can be prevented from repeating the same pattern, and the user can be prevented from being bored.

  Further, in the conversation processing device 1 and the conversation processing system 100 according to the first embodiment, it is determined whether or not there is a defect in natural language text extraction from the user's utterance, and when it is determined that there is a defect, the user again speaks. Since the determination unit 80 to be obtained is further provided, it is possible to prevent the generation of an output response sentence due to the failure of natural language text extraction or the malfunction of the conversation processing device 1 and the conversation processing system 100.

  Further, the conversation processing device 1 and the conversation processing system 100 according to the first embodiment include the input sentence storage unit 30 and the response sentence storage unit 40, and the output response sentence creation unit 60 is connected to the input sentence storage unit 30. While reading the input sentence, the response sentence storage unit 40 reads the response sentence for empathy, the response sentence for promoting confirmation, and the response sentence for specifying the conversation content, and creates an output response sentence. According to such a configuration, the information related to the input sentence for each user is stored in the input sentence storage unit 30, and a plurality of types of response sentences for empathy, confirmation prompting responses, and conversation content specifying response sentences are used. The response sentence can be stored in the response sentence storage unit 40. In addition, the constituent elements of the output response sentence can be read from each storage unit, and various output response sentences can be smoothly created according to the scene, the user's psychology, etc., and the conversation can be continued.

  Moreover, since the conversation processing device 1 and the conversation processing system 100 according to the first embodiment further include the input device 2 and the output device 4, the input form of the utterance from the user and the output of the output response sentence to the user The versatility of the conversation processing device 1 and the conversation processing system 100 can be enhanced by freely setting and changing the form.

  In addition, the conversation processing device 1 and the conversation processing system 100 according to the first embodiment acquire input-side environment information including the user, and the output response sentence creation unit 60 uses the environment information to accurately determine the user's psychology. Therefore, it is possible to create or select an appropriate output response sentence and to continue the conversation.

  Further, in the conversation processing device 1 and the conversation processing system 100 according to the first embodiment, the input sentence storage unit 30 stores the user's remarks at the time of past conversation as an already input sentence, and the output response sentence creation unit 60 includes: When it is determined that the input sentence matches the already input sentence or its tag, or is related to the purpose (aspiration) and the means (failure / reason), etc., the past response sentence indicating the past, the already input sentence, the confirmation prompting response The output response sentence including the sentence is created. According to such a configuration, the user talks with the previous conversation, highlights the user's inner face from both the past and current utterance contents, and the user's awareness based on both the past and current utterance contents. Can be encouraged.

[Second Embodiment]
Next, a second embodiment of the present invention will be described.

(Conversation processing device, conversation processing system)
FIG. 4 is a block diagram illustrating an example of the conversation processing system 110 according to the second embodiment. As shown in FIG. 4, the conversation processing system 110 includes at least a conversation processing device 1. Further, the conversation processing device 1 and the conversation processing system 110 according to the second embodiment have partially the same configuration as the conversation processing device 1 and the conversation processing system 100 according to the first embodiment. Therefore, in the following description, the conversation processing device 1 and the conversation processing system 110 according to the second embodiment will mainly be described with respect to the configuration different from the first embodiment, and the configuration common to the first embodiment will be the same. Reference numerals are assigned and explanations thereof are omitted.

As shown in FIG. 4, the conversation processing system 110 includes a robot apparatus 6 having an input unit 2X, an output unit 4X, and a robot control unit 5. The input unit 2X has the same function as the input device 2 described in the first embodiment, and includes a microphone and a speaker.
The output unit 4X has the same function as that of the output device 4 described in the first embodiment. However, in the second embodiment, the head (head) 121, the neck 125, and the trunk (trunk) that move according to the output response sentence. Part) 131, a hand (hand part) 132, and a leg (leg part) 133. As shown in FIG. 5, the head 121 has movable eyes (eye parts) 122, ears (ear parts) 123, nose (nose parts) 124, and mouth (mouth part) 126. As a simple structure, it has wrinkles, hair and the like. The conversation processing system 110 includes a servo meter (not shown) that moves the eyes 122, the ears 123, the nose 124, the eyelids, and the like. As an environment input unit 2Z that acquires environment information, a camera, an input button, a pulse meter, An input device (not shown) such as a thermometer and a line-of-sight detection device, and a display device (not shown) for outputting input voice data, pulse / pulse data, body temperature, and line-of-sight data are provided. That is, the conversation processing system 110 is a so-called robot type system. All of these may be included, but may be composed of only a part.

  The robot apparatus 6 further includes a robot input unit 141 that can input a user's speech other than voice, such as a touch panel or a keyboard, and a robot output unit 142 that can output the contents of an output response sentence as visual information. The robot output unit 142 can be configured by a display such as a liquid crystal display or a plasma display or a stereoscopic image display device, for example. FIG. 5 shows an example in which the robot input unit 141 and the robot output unit 142 are provided on a touch panel type device on the front surface of the body 131 (that is, the surface facing the user side), but the robot input unit 141 and the robot output unit are illustrated. 142 may be composed of different devices.

  The conversation processing system 110 is a robot control system software program, such as a real-time OS such as Raspberry Pi, Internet communication program, browser program, and JavaScript that mainly processes interface-related data such as input / output. Trademark) and the like. The main body (program) of the conversation processing device 1 may be arranged on the robot side, that is, inside the head 121 or the torso 131, or a main part other than the interface of the main body of the conversation processing device 1 may be arranged on the server. Good. Further, as the conversation processing device 1, a voice recognition server of Google (Google LLC) can be used. Non-language information such as pulse / pulse data, body temperature, line-of-sight data, etc. is transmitted to a non-language information recognizer (or server) (not shown) to identify the user's emotion. The non-linguistic information and emotion are transmitted to the output response sentence creating unit 60 together with the information of the natural language text identified from the user's voice data input from the input unit 2X and the information input from the environment input unit 2Z. Note that the voice recognition unit 12 and the voice generation unit 13 may be arranged inside the head 121 and the body 131 of the robot device 6 instead of the conversation processing device 1.

(Conversation processing method, program)
The conversation processing method of the second embodiment is basically the same as the conversation processing method of the first embodiment, and follows the procedure shown in the flowchart of FIG. Therefore, in the following description, the conversation processing method according to the second embodiment will be described mainly with respect to procedures and contents different from those of the first embodiment, and description of contents common to the first embodiment will be omitted.

  In step S <b> 11, the user can speak toward the input unit 2 </ b> X of the robot apparatus 6 and can input the content of the comment via the hand 132, an input device (not shown), and the robot input unit 141.

  In step S <b> 12, it is determined whether or not the output response sentence creation unit 60 has received the natural language text based on the voice uttered by the user using the output device having the text display function added to the robot device 6. it can.

In step S <b> 18, the output response text output from the output response text creation unit 60 is transmitted to the voice generation unit 13, converted into voice information, and transmitted to the robot control unit 5. Based on the received audio information, the robot control unit 5 emits a sound from the speaker provided on the mouth 126 of the head 121, and performs various movements such as reciprocity, whispering, tilting the head, gesturing gestures, and the like. 121, body 131, hand 132, and leg 133. In addition, the eyes 122, the ears 123, the nose 124, and the mouth 126 provided on the head 121 are operated so as to form a facial expression of emotions that gives the user peace of mind, gives a sense of trust, and makes them worried. The robot apparatus 6 performs non-language communication in accordance with the output response sentence to the user.
When the mouth 126 does not include a speaker and has a function of simply opening and closing, a sound may be emitted from a speaker or the like provided in the vicinity of the robot apparatus 6.

  For example, in the conversation example shown in FIG. 3, in the sentence OH51 of the agent, the head 121 crawls along with the voice output of “Fun, huh.” Are assembled in front of the conductor 131 or extended to the user. Further, in the sentence OH52, OH53,..., OH56 of the agent, it is expressed with eyes, mouth and hand gestures so as to express a gentle expression each time the “more details” output count is repeated. In addition, the agent's sentence OH57 expresses joyful eyes, mouth and hand movements along with a voice output of “It was good”. The output response sentence can be output as text via the robot output unit 142.

The body including the head 121, the torso 131, the hands 132, the legs 133, the eyes 122, the ears 123, the nose 124, the neck 125, the mouth 126, and the like included in the robot apparatus 6 may be physically present, The entire body or part of the body including the head 121, the torso 131, the hands 132, the legs 133, the eyes 122, the ears 123, the nose 124, the voice, the neck 125, the mouth 126, and the like may be virtually represented. For example, a user such as a two-dimensional moving image or still image, hologram, augmented reality, or virtual reality can be optically visually recognized as a character or avatar that imitates the shape of a robot or a person. Depending on the form, the head 121, the torso 131, the hand 132, the leg 133, and the like may be represented.
In this case, the input unit 2X and the output unit 4X are each configured by a microphone, a speaker, and the like provided in the vicinity of the robot device 6, but various movements associated with the sound are virtually transmitted in accordance with the sound output from the output unit 4X. The head 121, the torso 131, the hands 132, the legs 133, the eyes 122, the ears 123, the nose 124, the neck 125, and the mouth 126 are expressed in a realistic manner.
In this way, by virtually expressing at least a part of the body including the head 121, the torso 131, the hand 132, the leg 133, etc., even in an environment where the robot as an entity is not familiar, it is possible to communicate via the network. Can talk.
Further, the appearance of the robot apparatus 6 can be easily changed according to the user's preference so that the user has a sense of familiarity.

As described above, the conversation processing device 1 and the conversation processing system 110 according to the second embodiment basically have the same configuration as the conversation processing device 1 and the conversation processing system 100 of the first embodiment. The same effects as those of the first embodiment are obtained.
The conversation processing device 1 and the conversation processing system 110 according to the second embodiment include a head 121, a torso 131, a hand 132, and a leg 133 that move according to an output response sentence. The head 121 has an eye 122, an ear 123, a nose 124, a neck 125, and a mouth 126, and the conversation processing system 110 further includes an environment input unit 4Z. According to such a configuration, it is possible to realize a conversation processing system that is closer to the user, has various input forms and output forms, and can express a wide range of response contents.

  In addition, since the conversation processing system 110 according to the second embodiment includes the robot input unit 141 and the robot output unit 142, the conversation processing system 110 can approach the user via the body of the two-dimensional image and continue the conversation.

  Each configuration included in the conversation processing apparatus 1 described above has a computer system therein. Then, a program for realizing the functions of each component included in the above-described conversation processing device 1 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed. Thus, the processing in each configuration provided in the above-described conversation processing device 1 may be performed. Here, “loading and executing a program recorded on a recording medium into a computer system” includes installing the program in the computer system. The aforementioned “computer system” includes an OS and hardware such as peripheral devices.

  Further, the “computer system” may include a plurality of computer devices connected via a network including a communication line such as the Internet, WAN, LAN, and dedicated line. The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. As described above, the recording medium storing the program may be a non-transitory recording medium such as a CD-ROM.

  The recording medium also includes a recording medium provided inside or outside that is accessible from the distribution server in order to distribute the program. It should be noted that the program may be divided into a plurality of parts and downloaded at different timings, and the structure combined with each constituent provided in the conversation processing device 1 or the distribution server that distributes each of the divided programs may be different. Furthermore, the “computer-readable recording medium” holds a program for a certain period of time, such as a volatile memory (RAM) inside a computer system that becomes a server or a client when the program is transmitted via a network. Including things. The program may be for realizing a part of the functions described above. Further, the program may be a so-called difference file (difference program) that can realize the functions described above in combination with a program already recorded in the computer system.

  Moreover, you may implement | achieve part or all of the function mentioned above as integrated circuits, such as LSI (Large Scale Integration). Each function described above may be individually made into a processor, or a part or all of them may be integrated into a processor. Further, the method of circuit integration is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. In addition, when an integrated circuit technology that replaces LSI appears due to the advancement of semiconductor technology, an integrated circuit based on the technology may be used.

  Moreover, each structure with which the memory | storage device 3 mentioned above is provided may be comprised by the computer-readable recording medium, for example.

  The conversation processing device, the conversation processing system, the conversation processing method, and the program according to the embodiment to which the present invention is applied have been described above. The conversation processing apparatus, the conversation processing system, the conversation processing method, and the program according to the present invention have been described above. The present invention is not limited to the embodiments, and various modifications can be made within the scope of the gist of the present invention described in the claims.

  For example, as described above, the input form of the user's utterance content is not limited as long as the natural language text of the utterance can be extracted by the output response sentence creation unit 60. The user's utterance may be input by characters (text) using the input device 2 such as a keyboard or a digital pen instead of voice. Further, the output form of the output response sentence may not be the same as the input form, and is not particularly limited as long as the contents of the output response sentence can be transmitted to the user. The output response sentence may be output as text to the output device 4 such as a display instead of voice.

DESCRIPTION OF SYMBOLS 1 ... Conversation processing apparatus 2 ... Input device 4 ... Output device 30 ... Input word memory | storage part 40 ... Response sentence memory | storage part 60 ... Output response sentence preparation part 80 ... Judgment part (natural language element extraction defect judgment part)
100, 110 ... Conversation processing system

Claims (11)

Using at least part of the natural language elements included in the user's speech as input sentences,
An input sentence with a confirmation prompting response sentence in which a confirmation prompting response sentence for promoting confirmation of the content of the speech is connected to the end of the input sentence, a response sentence for empathy indicating empathy for the speech, and a conversation with the user What is claimed is: 1. A conversation processing apparatus comprising: an output response sentence creating unit that creates an output response sentence based on a response sentence including any one or more of response sentences for embodying conversation contents. The output response sentence creation unit creates the output response sentence by omitting any one or more of second-person expressions, modifiers, and predicates that indicate the user itself,
The output response sentence includes at least an input sentence with a confirmation prompting response sentence, and the conversation content instantiation response sentence,
The output response sentence creating unit creates the output response sentence including the input sentence with the confirmation prompting response sentence and the response sentence for instantiating the conversation content.
The conversation processing apparatus according to claim 1. A natural language element extraction failure determination unit that determines the presence or absence of extraction failure of the natural language element from the speech based on the input of the user and requests the user to speak again when it is determined that the failure exists Further comprising
The conversation processing apparatus according to claim 1 or 2. A conversation processing device according to any one of claims 1 to 3,
An input sentence storage unit that extracts the natural language element included in the statement and stores at least a part of the natural language element as the input sentence;
A log storage unit for recording past conversation contents of the user including the input sentence;
A response sentence storage unit in which the response sentence for empathy, the input sentence and the response sentence for promoting confirmation connected to the input sentence, and the response sentence for specifying the conversation content are stored;
With
The output response sentence creation unit reads the input sentence from the input sentence storage unit and the log storage unit, and connects to the response sentence for empathy, the input sentence and the input sentence from the response sentence storage unit A conversation processing system comprising: reading out a response response for promotion and a response message for specifying the conversation content, and creating the output response statement. An input device for inputting the user's speech;
An output device for outputting the output response sentence as audio information or visual information;
Further comprising
The conversation processing system according to claim 4. The input device acquires environment information on the input side including a user,
The conversation processing system according to claim 5, wherein the output response sentence creating unit creates the output response sentence based on the environment information. The input sentence storage unit and the log storage unit store the user's remarks at the time of past conversation as already input sentences,
The output response sentence creation unit, when it is determined that the input sentence matches or relates to the already input sentence or the tag of the already input sentence, the past response sentence indicating the past, the already input sentence, the confirmation promoting Creating the output response sentence including a response sentence;
The conversation processing system according to any one of claims 4 to 6. It further comprises any one or more of a head, a torso, a hand, and a leg that move according to the output response sentence.
The conversation processing system according to any one of claims 4 to 7. Any one or more of the head, the body part, the hand part, and the leg part are virtually represented,
The conversation processing system according to claim 8. An input sentence acquisition step of extracting a natural language element included in the user's utterance and acquiring an input sentence;
An input sentence with a confirmation prompting response sentence in which a confirmation prompting response sentence that promotes confirmation of the content of the comment is connected to the end of the input sentence, an empathy response sentence indicating empathy for the comment, and the user An output response sentence creating step for creating an output response sentence using a response sentence including any one or more of the conversation contents instantiation response sentences for embodying conversation contents;
Conversation processing method characterized by including. An input sentence acquisition step of extracting a natural language element included in the user's utterance and acquiring an input sentence;
An input sentence with a confirmation prompting response sentence in which a confirmation prompting response sentence that promotes confirmation of the content of the comment is connected to the end of the input sentence, an empathy response sentence indicating empathy for the comment, and the user An output response sentence creating step for creating an output response sentence using a response sentence including any one or more of the conversation contents instantiation response sentences for embodying conversation contents;
A program characterized by having executed.

Images