JP2007011674A - Reason explanation service processing method and apparatus using dialog robot, and program thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a mechanism for explaining to a user a reason why a service is executed in a robot system that autonomously executes a life support service.
An inference system 11 infers a user situation from environment information and behavior information stored in a distributed environment behavior DB 13, selects an executable service, causes the appliance 4 to execute the selected service, and executes the service. History information is accumulated in the service history DB 17. The dialogue control unit 23 of the dialogue robot 2 infers the context of the utterance information generated from the user's voice, and when the utterance is determined as the reason question, requests the reasoning system 11 to execute the reason question service. The inference system 11 identifies the question target service and generates an explanation of the execution reason based on the situation at the time of execution. The dialogue robot 2 utters the voice data generated from the explanation of the execution reason.
[Selection] Figure 2

Description

  The present invention is implemented in the life support robot system using a dialog robot that performs voice conversation with the user in cooperation with a life support robot system that autonomously controls the appliance to support the life of the user. The present invention relates to a reason explanation service processing method for explaining a reason for providing a service.

  With the background of ubiquitous environment, a life support robot system that is installed in a user's living space and autonomously supports the user's life has been realized. Life support robot systems based on a ubiquitous environment are often unconscious systems that do not require users to be particularly aware of the operation.

In order to maintain a comfortable environment and support user activities, the unconscious lifestyle support robot system infers the environment and user behavior based on various sensor information, and uses an appliance (communication function with a communication function). It provides various services for supporting life by controlling home appliances) and information terminal devices. For example, a life support robot system analyzes the user's movement / posture and room conditions based on sensor information and infers that the user is in a relaxed state. By playing music that is suitable for the environment or by maintaining the indoor temperature and humidity with an air conditioner, users can realize a comfortable living environment without consciously operating these appliances (see, non- Patent Document 1).
Mino Tetsuhiko, “Purpose and Outline of the Yukari Project—UKARI Project Report No. 1”, IPSJ 66th National Convention, 2004

  An unconscious life support robot system using a ubiquitous environment is convenient in that the user does not need to operate the appliance consciously. However, it is expected that the service provided by the life support robot system by autonomous control may not always be accepted by the user.

  The life support robot system will be able to support the whole of life, and the provided services will become more sophisticated and multifunctional, and the operating principle of the appliance will be complicated. For this reason, there is a possibility that it will not be possible to meet the individual requirements of the user with the autonomous control of the life support robot system alone.

  In addition, the mechanism of the life support robot system that is unconscious, that is, does not require conscious operation from the user, makes it difficult to recognize the existence of the system, which sometimes causes the user to feel weird. Can be. For example, a service that the life support robot system autonomously executes is comfortable for one user, but is uncomfortable for another user, and a service that is executed is rather different for the user. It is expected that a situation that seems to be annoying will occur. In such a case, the life support robot system eliminates such dissatisfaction and questions because the user wants to execute a different service or has a request to know the reason for the service execution. There is a need.

  Furthermore, there are cases where the user wants to know about matters other than the information stored in the life support robot system. For example, when the user himself / herself wants to operate the appliance, a situation may occur in which the sophisticated functions of each appliance cannot be fully used. In such a case, the user wants to know detailed information that is not covered by the life support robot system, so it is necessary to provide such information in the dialog robot.

  In this way, the unconscious lifestyle support robot system is equipped with a high level of situation handling capability as a universal user interface that can flexibly handle various situations, and to make the user feel friendly. It is necessary to provide a simple dialogue system.

  Therefore, as a user interface of the life support robot system, it is necessary for the user to support life support so that the user's ambiguous requests can be received more accurately and reflected in the control of the system, and the control status of the system can be understood more accurately. An interactive robot that can provide useful services and information for the user, and an interactive system using the interactive robot (see PCT / JP2004 / 002942).

  When such a life support robot system can provide various life support services, it is difficult for the user to grasp all the services to be executed, and why such services are provided. A mechanism is needed to explain to the user why.

  In addition, there is a need for a mechanism that can analyze how users evaluate the contents of services actually provided by the life support robot system and reflect them in the execution of subsequent services.

  The present invention has been made in view of the above-mentioned need, and its purpose is to explain the reason for the service provided by the life support robot system through the interactive robot when the user asks the reason for the execution. It is to provide a processing method that can be performed.

  Another object of the present invention is to ask the user the reason for service execution, explain the reason for the service execution, and feed back the user's evaluation of the executed service based on the user's reaction to the reason description. It is to provide a processing method that can be used.

  In order to autonomously control a plurality of appliances that execute services for supporting user's life, the present invention is based on sensor information measured in a predetermined space and the environment in the space and the user's behavior. An inference system that provides a service that controls the appliance by inferring the above, a distributed environment behavior database that accumulates environment information and behavior information as inference results by the inference system, and history information of services provided by the logical system A life support robot system having a service history information database to be accumulated, and a dialogue robot that cooperates with the life support robot system, and the context of the utterance is determined based on the utterance information generated by voice recognition processing of the voice data. Sound generated by inference and speech synthesis processing of speech information generated in the dialogue robot In a dialogue system including a dialogue robot that performs a voice dialogue with a user by outputting data, a speech explanation method for explaining the reason for execution of the service provided by the life support robot system is provided. The following processing steps are provided.

  The present invention includes 1) a process of inferring a user's situation from environmental information and action information stored in the distributed environment action database, and selecting a service to be provided to the user based on the user's situation; 2) A process of generating and transmitting control information for executing the selected service to the corresponding appliance, and causing the appliance to execute the service; and 3) including information regarding the status of the service at the time of execution of the service. A process of storing service history information in the service history information database; and 4) a process of converting the voice data into user utterance information by voice recognition processing when the interactive robot detects the voice data of the user. And 5) inferring the context of the user utterance information and executing the server on which the user utterance information is executed. And 6) a service history of the identified service based on the user utterance information, and a process of identifying the service that is the subject of the question, based on the user utterance information. A process of acquiring a service execution status from information and generating a description of the reason for executing the service based on the status; 7) voice data generated by voice synthesis processing of the reason description; And a processing process of uttering from.

  In addition, the present invention includes a processing process for detecting a user's reaction to the reason explanation when the interactive robot utters the voice data for the reason explanation, and the detected user reaction. And a process of learning user evaluation of the service for which the reason is explained and correcting the information accumulated in the distributed environment behavior database based on the learning result.

  Further, the present invention provides a process of determining whether or not user utterance information generated from voice data uttered by the user is a reason question within a predetermined period after execution of the service; If the user utterance information that is the reason question cannot be acquired within the period, a process for generating a reason question prompt for prompting the reason question in the dialogue robot, and a voice synthesis process for the reason question prompt is performed. And a process of uttering the voice data generated from the dialogue robot.

  The present invention is also the reason explanation service processing apparatus in which the dialogue robot utters voice data explaining the reason for executing the service provided by the life support robot system in the dialogue system, and 1) the distributed environment behavior Inferring the user's situation from environmental information and behavior information stored in the database, selecting a service to be provided to the user based on the user's situation, generating control information for executing the selected service, and responding Service determination means for transmitting to the appliance and causing the appliance to execute the service; and 2) service history information for storing service history information including information on the status of execution of the service in the service history information database. Storage means, and 3) the interactive robot When the user's voice data is detected, the voice data is subjected to voice recognition processing and converted into user utterance information, the context of the user utterance information is inferred, and the user utterance information is used to determine the reason for the service performed. And 4) a service specifying means for specifying the service subject to the question from the service history information database based on the user utterance information, and 4) a service from the service history information of the specified service. A dialog control means for acquiring a situation at the time of execution, generating a description of the reason for executing the service based on the status, and voice data generated by voice synthesis processing of the reason description from the dialog robot; It is to be prepared.

  The present invention also provides a reason explanation service processing program in which the dialogue robot utters voice data explaining the reason for executing the service provided by the life support robot system to the dialogue system, and 1) the distributed environment behavior A process of inferring a user's situation from environmental information and behavior information stored in a database, and selecting a service to be provided to the user based on the user's situation; 2) control information for executing the selected service And a process for causing the appliance to execute the service, and 3) for the service, service history information including information related to the status of execution of the service is stored in the service history information database. The accumulated process, and 4) the interactive robot is the user A process of recognizing the voice data and converting it into user utterance information when voice data is detected; 5) inferring the context of the user utterance information, and the user utterance information is A process of identifying the service that is the subject of the question from the service history information database based on the user utterance information when inferred that the reason question is, 6) from the service history information of the identified service A process of acquiring a service execution situation and generating a reason explanation of the service execution based on the situation; and 7) uttering speech data generated by speech synthesis processing of the reason explanation from the interactive robot This is a program for executing the processing process to be executed.

  According to the present invention, the reason for providing the service executed by the life support robot system can be explained to the user, so that the service is automatically provided without the user's request. The eerieness peculiar to a robot system can be eliminated.

  In addition, when a service different from the user's intention is executed, the reason for executing the service can be explained. Therefore, the cause of the inappropriate service being provided is shown to the user, and the user's dissatisfaction is reduced. be able to.

  In addition, since the user's response to the reason for executing the service can be obtained, the user's evaluation of the necessity and appropriateness of the executed service is analyzed based on the response, and the user's preference and service Can learn the necessity of

  FIG. 1 is a diagram showing a configuration example of a dialogue system in which the present invention is executed. In this embodiment, the dialogue system includes an unconscious life support robot system 1 applied to a family (residential space) composed of a plurality of users 3 (3a, 3b, 3c), a visible type Consists of a dialogue robot 2.

  The life support robot system 1 uses the ubiquitous environment, monitors the entire living space, executes a service by the home appliance (appliance) 4 having a communication function, and autonomously supports the life of the user 3.

  The dialogue robot 2 is installed in the living space of the user 3. The interactive robot 2 may have a function of moving autonomously and may be configured to freely move in the living space of the user 3. Although the dialogue robot 2 has only one entity as a system, it may have a plurality of housings equipped with processing units having the same configuration. The housing is sized so that it can be installed on a table, for example. When the interactive robot 2 has a plurality of housings, the processing means in the housing that detects the voice data among the interactive robots 2 installed in each room in the living space starts the processing operation as the interactive robot 2. . When the user 3 who is the conversation partner moves or the conversation with the user 3 who is in a remote place becomes necessary while the processing is continuing, the place closest to the conversation partner is set. The processing means of the casing takes over the processing, so that a single processing operation flow can be linked by a plurality of casings.

  This dialogue system is intended to provide a distributed collaboration between the unconscious lifestyle support robot system 1 and the visible dialogue robot 2. Such a distributed collaboration is a metaphor of the relationship between mother and child. Can be compared.

  The entire life space of the user 3 in which the unconscious life support robot system 1 is incorporated is referred to as a “mother metaphor”. In other words, it's like being in the house, watching over your family, appearing out of nowhere when you need it, and casually supporting your family. The dialogue robot 2 that handles the dialogue with the user 3 is referred to as a “child metaphor”. Although the conversation robot 2 does not have general social common sense, it can understand to some extent the preferences of specific users 3 who share a small living environment such as a home and the relationship between the users 3, and the family ( Users who want to be bitten by 3) and who are interested can be likened to being able to acquire advanced knowledge.

  In order to facilitate the relationship between the user 3 and the life support robot system 1, which is a mother-like presence for the user 3, the child metaphor dialogue robot 2 cooperates with the life support robot system 1 of the mother metaphor. It is considered to play the role of the “youngest child” who actively participates in the family and feels the presence of the mother.

  FIG. 2 is a diagram illustrating a configuration example of the life support robot system 1 and the dialogue robot 2.

  The life support robot system 1 includes an inference system 11, a sensor 12, a distributed environment behavior database 13, an event detection device 15, a service history database 17, and the like.

  The inference system 11 includes a knowledge database for inference. When the event detection device 15 detects an event, the inference system 11 acquires a context from the distributed environment behavior database 13 to obtain the environment in the living space, the operation of the user 3, the user 3 Analyzing the interaction (interaction), the interaction between the user 3 and the object, etc., inferring the action of the user 3 from the analysis result, and determining the service to be executed from the life support services implemented according to the conclusion , A processing system for controlling the corresponding appliance 4. Analysis results and inference results in the inference system 11 are accumulated in the distributed environment behavior database 13 as needed.

  The reasoning system 11 is implemented with a reason explanation service. The reason explanation service explains to the user 3 through the dialogue robot 2 the reason why the life support service is decided when the user 3 asks the reason for the execution of the life support service provided by the life support robot system 1. This is a process for performing a service.

  In addition, the inference system 11 performs feedback of user evaluation for the provided life support service. The user evaluation feedback is based on the user 3's reaction to the reason for service execution indicated by the reason explanation service. This is a process of reflecting the information unique to the user 3 in the distributed environment behavior database 13.

  The sensor 12 is a processing unit that has a communication function, measures and collects various data in the living space of the user 3, and transmits the collected sensor information to the distributed environment behavior database 13. The sensor 12 is, for example, a TV camera, a microphone, a floor sensor, an RFID tag monitor, an internal sensor of the appliance 4, or the like.

  The sensor information is data such as image data, sound data, pressure transition data, ID tag data, for example. Assume that an ID tag, for example, an RFID (Radio Frequency Identification) tag, in which identification information for non-contact radio wave system recognition is stored, is attached to an object owned by the user 3 or an object existing in the living space.

  The distributed environment behavior database 13 is a database system that accumulates and manages sensor information acquired from the sensor 12 and results analyzed or inferred by the inference system 11.

  FIG. 3 shows a configuration example of the distributed environment behavior database 13.

  The distributed environment behavior database 13 includes a distributed sensor information database 131, a distributed environment information database 132, a distributed operation information database 133, a distributed behavior information database 134, a person-object interaction database 135, a person-person interaction database 136, a specific information database 137, and the like. Consists of a database system.

  The distributed sensor information database 131 is a database that accumulates various types of sensor information transmitted from the sensor 12 at a predetermined time or opportunity.

  The distributed environment information database 132 is a database that accumulates environmental information such as the position, posture, temperature and humidity of the living space, and the internal state of the appliance 4.

For example, when the user 3x is reading a book toward the study desk, sensor information (time, ID data, position, image data, pressure data, internal sensor information of the service execution unit, etc.) measured by the sensor 12 is stored. It is analyzed by the inference system 11, and the following environmental information is generated and accumulated.
“Time, book ID data, position, status (held by hand),
Time, study desk lamp ID data, position, status (lit),
Time, study desk chair ID data, position, status (in use), ... "
The distributed motion information database 133 is a database that accumulates motion information indicating the position, posture, etc. of the user 3. For example, in the case of the above example, the sensor information is analyzed by the inference system 11 and the following operation information is accumulated.
"Time, user 3x ID data, position, posture (sitting position), ...".

The person-object interaction database 135 is a database for accumulating person-object interaction information. The person-object interaction information is information indicating a combination of the user 3 and the object in which the interaction occurs. For example, in the case of the above example, the inference system 11 analyzes the environment information and the operation information, and the following person-object interaction information is accumulated.
“People-things interaction 0001: user 3x, book, study desk, study desk chair, lamp,…”
The person-person interaction database 136 is a database that accumulates person-person interaction information. The person-person interaction information is information indicating a combination of the users 3 in which the interaction occurs. For example, it is assumed that a user 3a (father) and a user 3b (daughter) are sitting on a sofa and watching TV together. The inference system 11 analyzes the interaction from the motion information (position, posture, etc.) of the two people, and the following person-person interaction information is accumulated.
“People-person interaction 0011: user 3a, user 3b”.

The distributed behavior information database 134 is a database that accumulates behavior information indicating the behavior of the user 3. For example, in the case of the above-described example of the user 3x, the inference system 11 infers behavior information from environment information, operation information, person-object interaction information, person-person interaction information, etc., and accumulates the following behavior information. The
"User 3x: reading, studying ..."

  The unique information database 137 is a database that accumulates unique information indicating attributes for each user 3. The unique information is information inferred by the inference system 11 in addition to the physical characteristics and sex of the user 3 (for example, information on the preference of the user 3 and information on the necessity of the life support service).

  The event detection device 15 is a processing device that notifies the inference system 11 of event detection when an update of information in the distributed environment behavior database 13 or a service execution request from the interactive robot 2 is detected.

  The service history database 17 is a database that accumulates history information for each service executed by the inference system 11. The inference system 11 also refers to the information in the service history database 17 in the inference process. The service history information stored in the service history database 17 includes the name of the executed service, the service execution status, the service execution start time and end time, and the like. The service execution status is a part of information stored in the distributed environment behavior database 13 and is information related to the determination of the execution of the service, for example, information such as the position, operation, and behavior of the user 3. is there.

  The reasoning system 11, the distributed environment behavior database 13, the event detection device 15, and the service history database 17 of the life support robot system 1 can be implemented using known processing means or devices.

  The appliance 4 is a home appliance that has a data transmission / reception function and executes a predetermined service by the control from the inference system 11 or the operation of the user 3 itself.

  The dialogue robot 2 includes a voice recognition unit 21, a voice synthesis unit 22, a dialogue control unit 23, a dialogue strategy database 24, a situation information acquisition unit 25, a situation storage unit 26, a knowledge database 27, an execution request unit 29, and the like. .

  The voice recognition unit 21 is processing means for inputting voice data uttered by the user 3, recognizing the content of the input voice data, and converting it into the utterance information of the user 3.

  The voice synthesizer 22 is processing means for converting the robot utterance information generated by the dialogue controller 23 into voice data and uttering. The robot utterance information is data such as response sentence data for the conversation of the user 3 and an utterance sentence for asking the user 3.

  The dialogue control unit 23 analyzes the environment around the user 3 and the situation of the user 3 from the information stored in the situation storage unit 26, selects a service to be provided to the user 3 based on the analyzed situation, and dialogue strategy It is a processing means for performing a process of identifying a concept related to the situation of the user 3 based on the association information in the database 24 and generating robot utterance data using a language expression indicating the concept.

  The dialogue control unit 23 determines the execution of the reason explanation service based on the information in the distributed environment behavior database 13 and the service history database 17 stored in the situation storage unit 26 as a process for selecting a service.

  The dialogue strategy database 24 is a database in which the dialogue robot 2 infers a dialogue with the user 3 and accumulates information for generating a response for drawing the user 3 into his own dialogue or continuing.

  The situation information acquisition unit 25 acquires part of information such as environment information, operation information, action information, and unique information in the distributed environment action database 13 and part of service history information in the service history database 17 to obtain a situation storage unit 26. It is a processing means to store in.

  The knowledge database 27 is a database that accumulates knowledge information about the appliance 4 or a service executed on the appliance 4, for example, program information when the appliance 4 is a television, operation manual information when the appliance 4 is a microwave oven, and the like. is there.

  The execution request unit 29 is a processing unit that transmits the service execution request selected by the dialogue control unit 23 to the life support robot system 1. Further, the execution request unit 29 may transmit a service execution request directly to the appliance 4.

  In this example, the speech recognition unit 21, the speech synthesis unit 22, and the dialogue control unit 23 of the dialogue robot 2 are realized by an existing anthropomorphic spoken dialogue agent tool kit (Galatea Toolkit) (for example, http: //hil.tu -tokyo.ac.jp/~ galatea / galatea-jp.html, Prendinger, Helmut; Ishizuka, Mitsuru (Eds.); "Life-Like Characters Tools, Affective Functions, and Applications Series: Cognitive Technologies"; pp.187- 213; 2004; ISBN: 3-540-00867-5).

  In addition, the distributed environment behavior database 13 and the service history database 17 of the life support robot system 1 and the situation information acquisition unit 25 and the situation storage unit 26 of the dialogue robot 2 may be collectively configured as the distributed environment behavior database 13-1. it can.

  FIG. 4 is a diagram for explaining the outline of the processing of the present invention.

  Steps S1 to S4 show various life support service execution processes, Steps S5 to S9 show a reason explanation service execution process, and Step S10 shows a user evaluation feedback process.

  The life support robot system 1 acquires predetermined firing conditions (conditions that enable inference) for service execution based on the sensor information collected from the sensor 12 and the utterance information of the user 3 collected by the dialogue robot 2 (steps for enabling inference) S1). The inference system 11 determines the service content to be provided to the user 3 according to the context of the user 3 acquired from the distributed environment behavior database 13 or a context such as a dialogue (step S2). The corresponding appliance 4 is controlled to execute the determined service (step S3). Further, service execution history information on the determined service execution is stored in the service history database 17 (step S4).

  After that, for the service automatically executed by the life support robot system 1, the user 3 utters “why did you ○○” or “why? Suppose you asked.

  The dialogue robot 2 acquires the voice data “Why did you do it?” Of the user 3, analyzes it as a question sentence including a question expression such as “Why, why”, and the predetermined firing condition for executing the reason explanation service (The utterance of the user 3 is a question sentence) is acquired (step S5).

  The inference system 11 of the life support robot system 1 refers to the service history information in the service history database 17 and identifies the service for which the reason is requested from the currently executed services (step S6). The interactive robot 2 generates reason explanation text from information such as the service execution status of the identified service history information as robot utterance information (step S7), and speech synthesis of the robot utterance information (reason explanation text) Processing is performed, the user 3 is uttered, and a reason explanation service is executed (step S8).

  Furthermore, the life support robot system 1 stores the service history information of the executed reason explanation service in the service history database 17 (step S9).

  In addition, when the reason explanation service is executed in the process of step S8, if any response to the reason explanation can be obtained from the user 3, the inference system 11 uses the information on the reaction of the user 3 as a firing condition for user evaluation. Feedback is performed (step S10). The inference system 11 regards the response of the user 3 to the reason explanation as evaluation information for the service explained for the reason, and learns the user evaluation for the service concerning the reason explanation. Then, the learning result is reflected in the distributed environment behavior database 13.

  By such processing, even if the user 3 feels suspicious about the service automatically executed by the life support robot system 1 or is confused, it can be explained from the interactive robot 2 that is a child metaphor. The user 3 can understand the reason why the service is provided.

  Furthermore, since the user 3's reaction to the reason explanation can be regarded as an evaluation of the service for which the reason explanation is requested, feedback of the user evaluation can be performed, so that more appropriate service execution can be realized.

  In addition, when the life support robot system 1 executes a certain service for the first time, or when the user 3 does not ask for an explanation of the execution reason for the executed service, when the user evaluation feedback is not performed, You need to get a user rating for a running service. For this reason, the dialogue control unit 23 of the dialogue robot 2 refers to the dialogue strategy database 24 to generate the robot utterance information “Why say, say” and utter to the user 3. The reason for the reason for the dialog robot 2 is prompted, and the user 3 receives the utterance from the dialog robot 2 and answers “why?”, So that the process of step S5 is performed. A response can be expected and user evaluation can be obtained.

  FIG. 5 is a diagram showing an example of the processing flow of the dialogue system.

  The life support robot system 1 and the dialog robot 2 of the dialogue system cooperate with each other to repeatedly perform the life support service, reason explanation service, and user evaluation feedback processing.

Step S21: Speech data recognition processing The speech recognition unit 21 of the interactive robot 2 detects and inputs speech data uttered by the user 3, performs speech recognition processing on the speech data, and converts it into speech information.

Step S22: Status Information Acquisition Processing In parallel, the distributed environment behavior database 13 of the life support robot system 1 stores sensor information such as voice data and image data acquired from the sensor 12 in the distributed sensor information database 131, and sensor information Based on the information, the environment information in which the user 3 lives and the status information such as the state, operation, and behavior of the user 3 are acquired.

Step S23: Dialogue / situation inference / service selection processing When the voice data of the user 3 collected by the dialogue robot 2 and the sensor information collected by the sensor 12 are accumulated in the distributed environment behavior database 13, the event detection device 15 By referring to the distributed environment behavior database 13, an event such as addition or change of predetermined information is detected and notified to the inference system 11. The inference system 11 infers the situation of the user 3 using the utterance information and situation information of the user 3 related to the event detection as firing conditions, and selects an optimum service from the executable services.

  Alternatively, the dialogue control unit 23 of the dialogue robot 2 infers the context by analyzing the utterance information of the user 3, and infers the situation of the user 3 based on the inferred context and situation information. The optimum service may be selected from the above. In this case, the execution request unit 29 notifies the life support robot system 1 of the selected service execution request.

Step S24: Service Content Determination Process The inference system 11 determines the content of the selected life support service with reference to the information in the distributed environment behavior database 13.

Step S25: Execution Request Processing The inference system 11 generates control information for the appliance 4 to realize the selected service, and transmits it to the corresponding appliance 4.

Step S26: Speech synthesis processing When there is the robot utterance information of the interactive robot 2 as the control information of the service to be executed, the speech synthesizer 22 of the interactive robot 2 performs speech synthesis processing on the robot utterance information to obtain voice data. Is uttered to the user 3.

Step S27: Execution Content Storage Processing The inference system 11 stores the content of the executed service in the service history database 17 as service history information.

Step S28: Speech data recognition processing Thereafter, the user 3 utters “why did you XX” or “why △ △?” To the service automatically executed by the life support robot system 1 And ask the reason for service execution. The voice recognition unit 21 of the dialogue robot 2 detects and inputs voice data uttered by the user 3, and performs voice recognition processing on the input voice data to acquire utterance information.

Step S29: Dialogue / situation inference / service selection process The dialogue control unit 23 analyzes the utterance information of the user 3 to infer a context. Here, when the user utterance information includes a question expression such as “why, why” or an expression related to service execution, the utterance information asks the reason for service execution (reason question sentence). Infer that

  If acquisition of a reason question sentence is defined as a firing condition for executing the reason explanation service, the dialogue control unit 23 determines execution of the reason explanation service from the executable services. The execution request unit 29 notifies the life support robot system 1 of a request for executing the reason explanation service.

Step S30: Service identification process The inference system 11 refers to the service history information in the service history database 17 and identifies a service for which a reason is requested from among currently executed services.

Step S31: Reason Explanation Text Generation Processing The status information acquisition unit 25 acquires information such as the service execution status of the service history information for the specified service. Further, if necessary, the unique information of the user 3 is acquired from the distributed environment behavior database 13. The dialogue control unit 23 extracts an expression that is a key to explain the reason from the service execution state, generates a reason explanation, and uses it as robot utterance information.

Step S32: Speech Synthesis Processing The speech synthesis unit 22 performs speech synthesis processing on the reason explanation and utters the generated speech data to the user 3.

Step S33: Execution Content Storage Process The inference system 11 stores the reason execution service execution content in the service history database 17 as service history information.

Step S34: Speech data recognition processing After that, when the dialogue robot 2 utters an explanation of the reason, it is assumed that the user 3 responds such as "I understand", "Thank you", or "I don't need it".

  The voice recognition unit 21 detects and inputs the voice data “I understand” of the user 3, performs voice identification processing, and converts it into speech information.

Step S35: Dialogue / Situation Reasoning / Service Selection Process The dialogue control unit 23 analyzes the utterance information of the user 3 and infers the context. When the utterance information includes a positive expression such as “I understand”, it is inferred that the utterance information is a positive response sentence. Then, the user evaluation feedback process is selected. The execution request unit 29 notifies the life support robot system 1 of a user evaluation feedback request.

Step S36: Feedback Process The inference system 11 regards the response sentence of the user 3 to the reason explanation as user evaluation information for the service for which the reason explanation is requested, and learns the user evaluation for the service concerning the reason explanation. Then, based on the learning result, related information such as preference information in the unique information database 137 of the distributed environment behavior database 13 is corrected, or new information indicating the necessity of the service is added.

  For example, it learns from a positive response sentence such as “I understand” that the service for which the reason was explained was necessary and appropriate for the user 3. On the other hand, from the negative response sentence such as “I don't need it”, it is learned that the service whose reason has been explained is unnecessary or inappropriate for the user 3. Based on the learning result, the preference information of the user 3 in the unique information database 137 and the evaluation information about this service are corrected.

  In the user evaluation feedback process, in addition to the response sentence that is the utterance information of the user 3, sensor information such as a buzzing sound and a hitting sound of the dialog robot 2 and a gesture of the user 3 can be used.

  In addition, when the life support service of the life support robot system 1 is executed, if the user 3 already knows the reason for providing the service or the user 3 who has predicted predicts “thank you”, the reason explanation service is executed. May be omitted and the user evaluation feedback process may be performed immediately.

  Further, in the service specifying process in step S30, if the reason question sentence of the user utterance information includes information that can specify a service executed in the past, information that can specify a time zone, etc., service history information Identify the service for which the reason is sought by referring to the service execution time information.

  For example, if the user utterance information is “Why did you put on a weather forecast this morning?”, Based on the expression that can identify a certain time zone “This morning”, Identify the service for which the reason is sought.

  6 and 7 are diagrams showing a flow of processing in a specific example of the reason explanation service.

  FIG. 6 is a diagram showing the flow of processing of the reason explanation service when the life support service for automatically turning on the television switch is executed in the dialogue system of FIG.

  The life support robot system 1 detects a situation in which the users 3a and 3b are sitting in front of the appliance (television) 4 based on the sensor information collected from the cushion type sensor 12 in front of the television (step S41).

  The inference system 11 refers to the distributed environment behavior database 13 by notification of event detection, and uses the status information of the users 3a and 3b as a firing condition to execute the service “switch on TV” from among executable life support services. Determine (step S42).

  Then, “switch-on” control information is transmitted to the television which is the appliance 4. In addition, the dialogue robot 2 generates and speaks the robot utterance information “I turned on the TV” in order to inform the users 3a and 3b of the service contents (TV switch-on) performed by the life support robot system 1 (steps). S43).

  In addition, the life support robot system 1 stores service execution history information related to the provided service execution in the service history database 17. The service execution history information includes execution service (TV switch-on), service execution status (situation where users 3a and 3b are sitting in front of the TV), execution time information (execution start time, execution end time), etc. Is included (step S44).

  After that, since the television switch is automatically turned on by the life support robot system 1, it is assumed that the user 3a asks "Why did the television come on?"

  The dialogue control unit 23 of the dialogue robot 2 infers the utterance information as a reason question sentence because the utterance information extracted from the voice data of the user 3a by the voice recognition process includes the question expression “why”. The reasoning service is requested to be executed by the inference system 11 using the reason question sentence acquisition as the firing condition of the reason explanation service (step S45).

  The inference system 11 refers to the service history information in the service history database 17 and identifies the service “TV switch-on” for which the reason is sought from among the currently executed services (step S46).

  The dialogue control unit 23 of the dialogue robot 2 obtains the execution status of the service (situation where the users 3a and 3b are sitting in front of the television) from the service history information of the identified service, and the execution status and the like. From this, an expression such as “sitting in front of the TV” is extracted to generate a reason explanation “being sitting in front of the TV” as robot utterance information (step S47). Then, the voice synthesizer 22 utters the voice data generated by performing the voice synthesis process on the robot utterance information to the user 3a (step S48).

  Furthermore, the life support robot system 1 stores the service history information of the executed reason explanation service in the service history database 17 (step S49).

  In the process of step S48, if there is any response from the user 3a to the utterance of the interactive robot 2, the life support robot system 1 performs a feedback process of user evaluation using this as an ignition condition, and the result Is reflected in the information of the distributed environment behavior database 13 (step S410).

  FIG. 7 is a diagram showing the flow of processing of the reason explanation service when the life support service of attaching a recommended TV program according to the request of the user 3 is executed in the interactive system of FIG.

  The voice recognition unit 21 of the dialogue robot 2 detects voice data “Please turn on a recommended program” of the user 3a sitting on the sofa in front of the appliance (TV) 4 (step S51).

  The dialogue control unit 23 infers the context from “recommendation, program, attach” included in the utterance information of the user 3a generated by the speech recognition unit 21 through the speech recognition process, and executes the service from the executable services. Select “Turn on recommended TV program”. Furthermore, referring to the preference information of the user 3a in the distributed environment behavior database 13 acquired by the status information acquisition unit 25 and stored in the status storage unit 26, the program information currently being broadcast stored in the knowledge database 27, and the like. , Determine service contents. For example, if the user 3a's preference information stores information that “views many programs in which XX and △△ appear” is stored, the currently broadcast program in which XX or △△ appears from the program information. Search for. If the program a in which XX appears is being broadcast, the channel ch_n broadcasting the program a is acquired, and “turn on the TV and adjust the channel to ch_n” is set as the service content. (Step S52).

  Then, the life support robot system 1 transmits the control information “turn on the switch and set the channel to ch_n” to the television which is the appliance 4. Further, the dialogue robot 2 generates and utters the robot utterance information “I added program a” in order to make the user 3a confirm the executed service (step S53).

  Further, service execution history information related to the provided service execution is stored in the service history database 17. The service execution history information includes an execution service (switch is turned on and a channel is set to ch_n), a situation at the time of service execution (a situation where the user 3a is sitting in front of the TV, an utterance information of the user 3a "Recommended Information such as execution time information (execution start time, execution end time) is included (step S54).

  Thereafter, since the television is automatically turned on by the life support robot system 1 and the program a is attached, it is assumed that the user 3a asks "Why did you turn on this program?"

  The dialogue control unit 23 of the dialogue robot 2 infers the utterance information as a reason question sentence because the utterance information extracted from the voice data of the user 3a by the voice recognition processing includes the question expression “why”. The reasoning service is requested to be executed by the inference system 11 using the reason question sentence acquisition as the firing condition of the reason explanation service (step S55).

  The inference system 11 refers to the service history information in the service history database 17 and identifies the service “switch on and set channel to ch_n” for which the reason is requested from the currently executed services. (Step S56).

  From the service history information of the identified service, the dialogue control unit 23 of the dialogue robot 2 determines the situation at the time of execution of the service (the situation in which the users 3a and 3b are sitting in front of the TV, the utterance information of the user 3a Turn on the program and get ")". Furthermore, preference information (viewing many programs in which XX and △ Δ appear) among the unique information of the user 3a in the distributed environment behavior database 13, program information currently being broadcast, and the like are acquired. Then, based on these pieces of information, a reason explanation sentence “because XX is out” is generated and used as robot utterance information (step S57).

  Then, the voice synthesizer 22 utters the voice data generated by performing the voice synthesis process on the robot utterance information to the user 3a (step S58).

  Furthermore, the life support robot system 1 stores the service history information of the executed reason explanation service in the service history database 17 (step S59).

  In the process of step S58, if there is any response from the user 3a to the utterance of the interactive robot 2, the life support robot system 1 performs a feedback process of user evaluation using this as an ignition condition, and the result Is reflected in the information of the distributed environment behavior database 13 (step S510).

  While the present invention has been described above with reference to the embodiments, it is obvious that the present invention can be variously modified within the scope of the gist thereof.

  The present invention can also be implemented as a processing program that is read and executed by a computer. The processing program for realizing the present invention can be stored in an appropriate recording medium such as a portable medium memory, a semiconductor memory, and a hard disk that can be read by a computer. It is provided by transmission / reception using various communication networks via a communication interface.

It is a figure which shows the structural example in implementation of the dialogue system of this invention. It is a figure which shows the structural example of a life support robot system and a dialogue robot. It is a figure which shows the structural example of a distributed environment action database. It is a figure for demonstrating the outline | summary of the process of this invention. It is a figure which shows the example of the flow of a process of a dialogue system. It is a figure which shows the flow of a process in the specific example of a reason explanation service. It is a figure which shows the flow of a process in the specific example of a reason explanation service.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Life support robot system 11 Reasoning system 12 Sensor 13 Distributed environment action database 15 Event detection apparatus 17 Service history database 2 Dialogue robot 21 Speech recognition part 22 Speech synthesizer 23 Dialogue control part 24 Dialogue strategy database 25 Situation information acquisition part 26 Situation memory Part 27 Knowledge database 29 Execution request part 3 User 4 Appliance

Claims (9)

In order to autonomously control a plurality of appliances that perform services for supporting users' lives, the environment in the space and user behavior are inferred based on sensor information measured in a predetermined space. An inference system for providing a service for controlling the appliance, a distributed environment action database for storing environment information and action information as inference results by the inference system, and a service history for storing history information of services provided by the logical system A life support robot system including an information database, and a dialog robot that cooperates with the life support robot system, inferring the context of the utterance based on the utterance information generated by voice recognition processing of voice data, and dialog Speech data generated by speech synthesis processing of speech information generated in the robot By force, in an interactive system with an interactive robot for users and voice interaction,
A reason explanation service processing method in which the interactive robot speaks voice data explaining a reason for executing a service provided by the life support robot system,
A process of inferring a user's situation from environmental information and action information stored in the distributed environment action database, and selecting a service to be provided to the user based on the user's situation;
A process of generating control information for executing the selected service and transmitting it to the corresponding appliance, causing the appliance to execute the service;
For the service, a process of storing service history information including information on the status of execution of the service in the service history information database;
A process of converting the voice data into user utterance information by performing voice recognition processing when the interactive robot detects the voice data of the user;
When the context of the user utterance information is inferred and the user utterance information is inferred to be a reason for a service executed, based on the user utterance information, the target of the question is retrieved from the service history information database. A process that identifies the service
A process of obtaining a service execution status from the service history information of the identified service and generating a description of the reason for the service execution based on the status;
A reason explanation service processing method using a dialog robot, comprising: a process of uttering voice data generated by voice synthesis processing of the reason explanation from the dialog robot. A process of detecting a user response to the reason explanation when the dialogue robot utters the voice data of the reason explanation;
Using the detected user's reaction, learning a user evaluation of the service for which the reason is explained, and a process of correcting information accumulated in the distributed environment behavior database based on the learning result. The reason explanation service processing method using the dialog robot according to claim 1. A process of determining whether user utterance information generated from voice data uttered by the user is a reason question within a predetermined period after execution of the service;
A process of generating a reason question prompt for prompting the reason question in the dialogue robot when the user utterance information that is the reason question cannot be acquired within the predetermined period;
The reason explanation service processing method using the interactive robot according to claim 1, further comprising: a process of uttering voice data generated by voice synthesis processing of the reason question prompting from the dialog robot. In order to autonomously control a plurality of appliances that perform services for supporting users' lives, the environment in the space and user behavior are inferred based on sensor information measured in a predetermined space. An inference system for providing a service for controlling the appliance, a distributed environment action database for storing environment information and action information as inference results by the inference system, and a service history for storing history information of services provided by the logical system A life support robot system including an information database, and a dialog robot that cooperates with the life support robot system, inferring the context of the utterance based on the utterance information generated by voice recognition processing of voice data, and dialog Speech data generated by speech synthesis processing of speech information generated in the robot By force, in an interactive system with an interactive robot for users and voice interaction,
A reason explanation service processing device for the conversation robot to utter speech data explaining the reason for execution of the service provided by the life support robot system,
Inferring the user's situation from the environmental information and behavior information stored in the distributed environmental behavior database, selecting a service to be provided to the user based on the user's situation, and generating control information for executing the selected service A service determination means for transmitting to the corresponding appliance and causing the appliance to execute the service;
Service history information storage means for storing, in the service history information database, service history information including information related to the service execution status for the service;
When the interactive robot detects the voice data of the user, the voice data is subjected to voice recognition processing and converted into user utterance information, the context of the user utterance information is inferred, and the user utterance information is executed. A service specifying means for specifying a service that is a target of the question from the service history information database based on the user utterance information when inferring that the question is a service reason question;
Obtaining a service execution status from the service history information of the identified service, generating a description of the reason for executing the service based on the status, and generating voice data generated by performing speech synthesis on the reason description, A reason explanation service processing apparatus using a dialog robot, comprising: a dialog control means for speaking from the dialog robot. When the dialogue robot utters the voice data of the reason explanation, the user's reaction to the reason explanation is detected, and the user evaluation of the service that is the object of the reason explanation is performed using the detected user reaction. The reason explanation service processing apparatus using the interactive robot according to claim 4, further comprising user evaluation feedback means for learning and correcting information accumulated in the distributed environment behavior database based on the learning result. The dialogue control means determines whether the user utterance information generated from the voice data uttered by the user is a reason question within a predetermined period after execution of the service, and within the predetermined period If the user utterance information that is the reason question cannot be acquired, the reason question prompting for prompting the reason question in the dialogue robot is generated, and the voice data generated by voice synthesis processing of the reason question prompt is generated. 5. The reason explanation service processing apparatus using a dialog robot according to claim 4, wherein the user speaks from the dialog robot. In order to autonomously control a plurality of appliances that perform services for supporting users' lives, the environment in the space and user behavior are inferred based on sensor information measured in a predetermined space. An inference system for providing a service for controlling the appliance, a distributed environment action database for storing environment information and action information as inference results by the inference system, and a service history for storing history information of services provided by the logical system A life support robot system including an information database, and a dialog robot that cooperates with the life support robot system, inferring the context of the utterance based on the utterance information generated by voice recognition processing of voice data, and dialog Speech data generated by speech synthesis processing of speech information generated in the robot By force, in an interactive system with an interactive robot for users and voice interaction,
A reason explanation service processing program in which the dialogue robot utters voice data explaining the reason for execution of the service provided by the life support robot system to the dialogue system,
A process of inferring a user's situation from environmental information and action information stored in the distributed environment action database, and selecting a service to be provided to the user based on the user's situation;
A process of generating control information for executing the selected service and transmitting it to the corresponding appliance, causing the appliance to execute the service;
For the service, a process of storing service history information including information on the status of execution of the service in the service history information database;
A process of converting the voice data into user utterance information by performing voice recognition processing when the interactive robot detects the voice data of the user;
When the context of the user utterance information is inferred and the user utterance information is inferred to be a reason question of the executed service, the target of the question is retrieved from the service history information database based on the user utterance information. The process of identifying the service to be
A process of obtaining a service execution status from the service history information of the identified service and generating a description of the reason for the service execution based on the status;
A reason explanation service processing program using a dialogue robot for executing voice data generated by voice synthesis processing of the reason explanation from the dialogue robot. In the dialogue system,
A process of detecting a user response to the reason explanation when the dialogue robot utters the voice data of the reason explanation;
A process of learning user evaluation of the service for which the reason is explained using the detected user reaction, and correcting information accumulated in the distributed environment behavior database based on the learning result is executed. A reason explanation service processing program using the dialogue robot according to claim 7. In the dialogue system,
A process of determining whether user utterance information generated from voice data uttered by the user is a reason question within a predetermined period after execution of the service;
A process of generating a reason question prompt for prompting the reason question in the dialogue robot when the user utterance information that is the reason question cannot be acquired within the predetermined period;
The reason explanation service processing program using a dialog robot according to claim 7, wherein the voice data generated by voice synthesis processing of the reason question prompting is executed by a process of speaking from the dialog robot.

Images