WO2024014383A1 - Dispositif de traitement d'informations, procédé de traitement d'informations, dispositif terminal et programme de terminal - Google Patents

Dispositif de traitement d'informations, procédé de traitement d'informations, dispositif terminal et programme de terminal Download PDF

Info

Publication number
WO2024014383A1
WO2024014383A1 PCT/JP2023/025063 JP2023025063W WO2024014383A1 WO 2024014383 A1 WO2024014383 A1 WO 2024014383A1 JP 2023025063 W JP2023025063 W JP 2023025063W WO 2024014383 A1 WO2024014383 A1 WO 2024014383A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
question
question information
user
answer
Prior art date
Application number
PCT/JP2023/025063
Other languages
English (en)
Japanese (ja)
Inventor
思瑶 呉越
裕士 瀧本
正弘 高橋
Original Assignee
ソニーグループ株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by ソニーグループ株式会社 filed Critical ソニーグループ株式会社
Publication of WO2024014383A1 publication Critical patent/WO2024014383A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/903—Querying
    • G06F16/9035—Filtering based on additional data, e.g. user or group profiles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Definitions

  • the present invention relates to an information processing device, an information processing method, a terminal device, and a terminal program.
  • the present disclosure proposes an information processing device, an information processing method, a terminal device, and a terminal program that can improve the efficiency of interaction with users.
  • an information processing device acquires a series of question candidates that can be presented to a requester who has requested a dialogue, from among question information registered in a knowledge database. omitting presentation to the requester of the question information included in the series of question candidates based on the interaction history between the acquisition unit and the user realized by providing the question information to the user; and a determining unit that determines question information to be omitted.
  • FIG. 1 is a diagram illustrating an example of a system according to an embodiment.
  • FIG. 2 is a diagram illustrating an example of the knowledge database according to the embodiment.
  • FIG. 3 is an explanatory diagram illustrating an overview of information processing according to the embodiment.
  • FIG. 4 is a diagram illustrating a configuration example of an information processing device according to an embodiment.
  • FIG. 5 is a diagram illustrating a configuration example of a terminal device according to an embodiment.
  • FIG. 6 is a flowchart showing an accumulation processing procedure for accumulating interaction history.
  • FIG. 7 is a flowchart (1) showing the procedure for updating the knowledge graph.
  • FIG. 8 is a diagram (1) showing an example of a knowledge graph updating method.
  • FIG. 9 is a flowchart (2) showing the knowledge graph update processing procedure.
  • FIG. 10 is a diagram (2) illustrating an example of a knowledge graph updating method.
  • FIG. 11 is a flowchart showing a determination processing procedure for determining question information to be omitted.
  • FIG. 12 is a diagram showing an example of a display screen when question information is displayed.
  • FIG. 13 is a diagram illustrating an example of a situation in which question information to be omitted is presented.
  • FIG. 14 is a block diagram illustrating an example of a hardware configuration of a computer corresponding to an apparatus according to an embodiment and a modification of the present disclosure.
  • the information processing device uses a knowledge graph obtained by extracting knowledge from a document to generate question candidates for the content to be asked back.
  • the information processing device feeds back judgment information for determining whether to present a question candidate to the user to the knowledge graph based on the analysis result of the dialogue history.
  • the feedback results can be reflected in the next inquiry, so that, for example, presentation of question candidates that are obvious to the user making the inquiry can be omitted.
  • the information processing device can improve the efficiency of interaction with the user.
  • FIG. 1 is a diagram illustrating an example of a system according to an embodiment.
  • FIG. 1 shows an information processing system 1 as an example of a system according to an embodiment.
  • the information processing system 1 may include a terminal device 10 and an information processing device 100. Further, the terminal device 10 and the information processing device 100 may be communicably connected via the network N by wire or wirelessly.
  • the information processing device 100 is an example of an information processing device according to the proposed technology of the present disclosure, and performs information processing according to the embodiment. Further, for example, the information processing apparatus 100 corresponds to a dialogue support system DL that realizes dialogue with a user who has inquired about a predetermined service (hereinafter referred to as "service SA").
  • service SA a predetermined service
  • Examples of services SA to which the information processing device 100 can be applied include “life insurance benefit payment service” and “rental property search service.”
  • the service SA is not limited to such an example.
  • the service SA will be described as a "life insurance benefit payment service.”
  • the terminal device 10 may be used by the user when accessing the dialogue support system DL in order to enjoy the service SA.
  • the terminal device 10 may be a smartphone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), or the like.
  • predetermined application software that realizes transmission and reception of information between the terminal device 10 and the information processing device 100 may be installed.
  • application software may be general-purpose application software such as a web browser, or may be implemented as dedicated application software corresponding to the dialogue support system DL.
  • the terminal device 10 can be said to be an edge computer that performs edge processing near the user.
  • the information processing device 100 may be, for example, a cloud computer that performs processing on the cloud side, that is, a server device.
  • FIG. 2 is a diagram illustrating an example of the knowledge database according to the embodiment.
  • the data structure in a knowledge database may be defined by a knowledge graph.
  • a knowledge graph may be generated for each type of benefit handled by service SA (life insurance benefit payment service), and FIG. 2 shows an example of knowledge graph KG1 corresponding to disaster hospitalization benefit.
  • SA life insurance benefit payment service
  • the knowledge graph KG1 may have a graph structure that describes knowledge extracted for each item in terms and conditions information corresponding to disaster hospitalization benefits out of the terms and conditions information DA121 related to service SA. .
  • Knowledge Graph KG1 converts related knowledge information into a graph structure among the knowledge information written for each item such as "type”, “reason for payment”, “amount of payment”, and “reason for exemption”. It may be generated by connecting it so as to drop it.
  • knowledge graph KG1 includes knowledge information on item AG1 (reason for payment), knowledge information on item AG2 (payment amount), and knowledge information on item AG3 (reason for exemption) for the item "disaster hospitalization benefit".
  • the knowledge information in item AG4 (AND condition) and the knowledge information in item AG5 (guarantee content) may be generated in a form that links them.
  • the information processing device 100 extracts items that are not directly described in the terms and conditions information DA121 based on the terms and conditions information DA121, and applies knowledge information to the extracted items to create the knowledge graph KG1. You can improve the accuracy.
  • item AG4 and item AG5 are not directly described in the terms and conditions information DA121.
  • the information processing device 100 extracts the item AG4 and the item AG5 based on the terms and conditions information DA121, and generates the knowledge graph KG1 by applying knowledge information to these items.
  • the information processing device 100 may obtain knowledge information for each item by understanding the meaning of the terms and conditions information DA121 using natural language processing, and may add the obtained information when generating the knowledge graph KG1.
  • condition information CO1, condition information CO2, condition information CO3, etc. are added to the knowledge graph KG1 as knowledge information corresponding to item AG4 (AND condition). These condition information are used to determine the official answer to the question input by the user, and are used as question information to ask the user what information is missing from the user's question. That is, in the example shown in FIG. 2, the knowledge graph KG1 is associated with question information Q11, question information Q12, question information Q13, etc. used for asking back questions in response to question-type input information input by the user. It can be said that it is registered in the database. In the following, an example of information processing according to the embodiment will be described with the knowledge database 122 having the knowledge graph KG1.
  • FIG. 3 is an explanatory diagram illustrating an overview of information processing according to the embodiment.
  • Information processing according to the embodiment is realized between the terminal device 10 and the information processing device 100 in the information processing system 1.
  • FIG. 3 shows, as an example, a scene in which a user Ux, who is an arbitrary user, interacts with a chatbot by utilizing the dialogue support system DL in order to enjoy the service SA.
  • the user Ux uses the terminal device 10 to input the following information: "I was hospitalized due to a traffic accident. Will I receive benefits?" and sends the message to the dialogue support system DL. I am inquiring.
  • the terminal device 10 transmits input information IN11 in the form of a question input by the user Ux to the information processing device 100.
  • the user Ux can be said to be the requester who requested the interaction.
  • the information processing device 100 When the information processing device 100 receives the input information IN11 from the terminal device 10 (step S11), it executes an analysis process on the input information IN11 (step S12). Specifically, the information processing device 100 performs syntax analysis and semantic analysis on the input information IN11.
  • the information processing device 100 refers to the knowledge database 122 and extracts a series of question candidates corresponding to the results of the analysis process (step S13). For example, the information processing device 100 refers to the knowledge graph KG1 and identifies a plurality of pieces of condition information according to the question content indicated by the input information IN11. Then, the information processing apparatus 100 extracts question information that asks whether or not the specified condition information is satisfied, for each piece of specified condition information. The information processing apparatus 100 then determines the extracted question information group as a series of question candidates to be presented to the user Ux.
  • the information processing device 100 displays question information Q11 with the content “Are you hospitalized at a hospital or clinic?” and question information Q12 with the content “Are you hospitalized for 5 days or more?”
  • question information Q11 and Q12 can be information for asking questions in response to the content of the question indicated by the input information IN11.
  • the information processing device 100 executes unnecessary information filter processing based on the interaction history (step S14). Specifically, the information processing apparatus 100 determines question information to be omitted from being presented to the user Ux, from among the question information Q11 and Q12 included in the series of question candidates, based on the dialogue history. For example, the information processing device 100 determines which of the question information Q11 and Q12 is obvious to the user Ux (question information for which it is possible to predict what kind of answer the user Ux will give) based on the dialogue history. , the question information to be omitted is determined according to the determination result.
  • the dialogue history here may be history information of dialogues realized with an unspecified number of users using various types of question information registered in the knowledge database 122.
  • the knowledge database 122 is updated based on the interaction history. For example, meta information for determining whether to present the question information to the user may be added to the question information registered as the knowledge graph KG1 based on the analysis result of the dialogue history. Therefore, the information processing apparatus 100 essentially refers to the knowledge graph KG1 that has received feedback based on the analysis results of the dialogue history, and determines the question information to be omitted.
  • the information processing device 100 generates question information to be actually presented to the user Ux (step S15). For example, it is assumed that the information processing apparatus 100 determines the question information Q11 among the question information Q11 and Q12 as the question information to be omitted in step S14. In this case, the information processing device 100 generates question information to be actually presented based on the question information Q12 remaining after excluding the question information Q11. For example, the information processing device 100 may generate the question information Q12 to be presented by performing arbitrary processing on the question information Q12, such as adding missing words.
  • the information processing device 100 presents the question information Q12 to be presented to the user Ux (step S16). Specifically, the information processing device 100 performs output control so that the question information Q12 to be presented is output from an output unit (for example, a display screen, a speaker, etc.) included in the terminal device 10.
  • an output unit for example, a display screen, a speaker, etc.
  • the terminal device 10 displays on the display screen the question information Q12 that says "Are you hospitalized for 5 days or more?” in accordance with the output control by the information processing device 100. . That is, in the example of FIG. 3, the chatbot of the dialogue support system DL skips presenting the question information Q11 to the user Ux. On the other hand, the chatbot replies to the input information IN11 in the form of a question by the user Ux with the content, "Are you hospitalized for more than 5 consecutive days?"
  • the information processing device 100 executes an accumulation process of accumulating a dialogue history based on the answer (step S17).
  • the information processing device 100 executes an update process to update the knowledge database 122 based on the interaction history (step S18).
  • the interaction history is analyzed, and based on the analysis results, feedback is provided that meta information for determining whether or not to present question information to the user is linked to the knowledge graph KG1. It will be done.
  • the information processing device 100 selects omitted question information that does not need to be presented because it is obvious what kind of answer the user Ux will answer, out of a series of candidate question information to be presented to the user Ux. Determined based on dialogue history. Then, the information processing apparatus 100 presents only the question information other than the question information to be omitted to the user Ux. As a result, the information processing device 100 can improve the efficiency of interaction with the user Ux.
  • FIG. 4 is a diagram illustrating a configuration example of the information processing device 100 according to the embodiment.
  • the information processing device 100 includes a communication section 110, a storage section 120, and a control section 130.
  • the communication unit 110 is realized by, for example, a NIC (Network Interface Card).
  • the communication unit 110 is wirelessly connected to the network N, and transmits and receives information to and from the terminal device 10, for example.
  • the storage unit 120 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk.
  • the storage unit 120 includes a terms and conditions information database 121 , a knowledge database 122 , a user information database 123 , and a dialogue history database 124 .
  • the terms and conditions information database 121 may store terms and conditions information corresponding to the service SA.
  • the terms and conditions information database 121 may store a terms and conditions table in which knowledge information is assigned to each item related to the service SA, as described with reference to FIG.
  • the knowledge database 122 may store a knowledge graph generated from the terms and conditions information corresponding to the service SA. Further, information related as a knowledge graph (for example, condition information) may be used as question information provided to a user who has requested a dialogue.
  • information related as a knowledge graph for example, condition information
  • question information may be used as question information provided to a user who has requested a dialogue.
  • the user information database 123 may store various attribute information indicating the user's gender, age, etc., as information regarding the user.
  • the user information database 123 may store attribute information of users who have a history of dialogue using the dialogue support system DL.
  • the user information database 123 may include items such as "user ID,” "attribute information,” and "service SA type.”
  • the dialogue history database 124 may store history information of dialogues realized with users.
  • the dialogue history database 124 may include items such as "user ID,” “attribute information,” “question information,” and “answer content.”
  • the control unit 130 is realized by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like executing various programs stored in a storage device inside the information processing device 100 using the RAM as a work area. Further, the control unit 130 is realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).
  • ASIC Application Specific Integrated Circuit
  • FPGA Field Programmable Gate Array
  • the control unit 130 includes a reception unit 131, an analysis unit 132, an extraction unit 133, a filter processing unit 134, a generation unit 135, an output control unit 136, an accumulation unit 137, and a detection unit 132. It has a section 138 and an update section 139, and realizes or executes information processing functions and operations described below.
  • the internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 4, and may be any other configuration as long as it performs information processing to be described later.
  • the connection relationship between the respective processing units included in the control unit 130 is not limited to the connection relationship shown in FIG. 4, and may be other connection relationships.
  • the reception unit 131 receives a request for dialogue from a user. For example, the accepting unit 131 accepts input information regarding the contents of an inquiry regarding the service SA. More specifically, the reception unit 131 receives input information from the user's terminal device 10.
  • the analysis unit 132 executes an analysis process on the received input information. For example, the analysis unit 132 may estimate characteristic slots by performing syntax analysis and semantic analysis on the input information.
  • the extraction unit 133 is a processing unit that corresponds to an acquisition unit.
  • the extraction unit 133 extracts (obtains) a series of question candidates that can be presented to the user who has requested a dialogue, from among the question information registered in the knowledge database 122.
  • the extraction unit 133 extracts a series of question candidates that may be presented to a user who has input a question regarding the service SA from among question information registered as a knowledge graph. More specifically, the extraction unit 133 may extract a series of question sentences for asking the user whether or not the condition information corresponding to the user's question is satisfied as a series of question candidates.
  • the filter processing section 134 is a processing section corresponding to a determining section. Based on the dialogue history, the filter processing unit 134 determines, from among the question information included in the series of question candidates, question information to be omitted from being presented to the user who requested the dialogue. For example, the filter processing unit 134 determines that, based on trends obtained from the interaction history, there is question information in a series of question candidates that allows prediction of the answers the user will give. In this case, the question information is determined as the question information to be omitted.
  • the generation unit 135 generates question information to be presented based on question information included in the series of question candidates, excluding question information to be omitted.
  • the output control unit 136 presents the question information to be presented to the requesting user. Specifically, the output control unit 136 performs output control so that the question information to be presented is output from an output unit (for example, a display screen, a speaker, etc.) included in the terminal device 10.
  • an output unit for example, a display screen, a speaker, etc.
  • the storage unit 137 executes a storage process of storing a dialogue history based on the content of the answer.
  • the detection unit 138 detects a tendency regarding the content of answers given by the user to the question information based on the dialogue history.
  • the update unit 139 updates the knowledge database 122 based on the detected trends.
  • the detection unit 138 detects the response tendency of the user based on the conversation history, i.e., what content the user tends to answer in response to the question information.
  • the updating unit 139 updates the knowledge database 122 based on the response tendency.
  • the detection unit 138 may classify users who have responded to the question information by response content, and may detect response trends based on the number of users for each response content.
  • the updating unit 139 links answer information indicating the detected answer tendency to the question information. It may be registered in the knowledge database 122 in the state.
  • the filter processing unit 134 determines whether or not there is question information associated with answer information in the question information included in the series of question candidates in the knowledge database 122. . If the filter processing unit 134 determines that there is question information with which answer information is linked, it predicts that the user will answer the question information with the answer content indicated by the answer information, and The information is determined as question information to be omitted.
  • the detection unit 138 may also detect attribute trends, based on the interaction history, to determine what kind of attributes the users who answered with the response content indicated by the response trends tend to have.
  • the update unit 139 updates the knowledge database 122 based on the attribute tendency.
  • the detection unit 138 classifies the users who responded with the response content indicated by the response tendency by predetermined attributes, and detects the attribute tendency based on the number of users for each attribute. Furthermore, if it is detected that a user who belongs to one of the attributes used for classification tends to answer with the answer content indicated by the answer tendency, the updating unit 139 updates the detected attribute. Attribute information indicating trends is registered in a knowledge database in a state where it is linked to question information.
  • the filter processing unit 134 determines whether or not there is question information associated with attribute information among the question information included in the series of question candidates in the knowledge database 122. . Then, if there is question information linked with attribute information, the filter processing unit 134 determines the question information based on the similarity between the linked attribute information and the attribute information possessed by the requesting user. It is determined whether the question information is to be omitted. For example, if the filter processing unit 134 determines that there is similarity between the attribute information linked to the question information and the attribute information possessed by the requesting user, the filter processing unit 134 selects the question information to be omitted. It may be determined as question information.
  • FIG. 5 is a diagram showing a configuration example of the terminal device 10 according to the embodiment.
  • the terminal device 10 includes a communication section 11, a storage section 12, a display section 13, an operation section 14, and a control section 15.
  • the communication unit 11 is realized by, for example, a NIC or the like.
  • the communication unit 11 is wirelessly connected to the network N, and transmits and receives information to and from the information processing device 100, for example.
  • the storage unit 12 is realized by, for example, a semiconductor memory device such as a RAM or a flash memory, or a storage device such as a hard disk or an optical disk.
  • the storage unit 12 may store, for example, information regarding applications installed in the terminal device 10 (for example, the terminal program according to the embodiment).
  • the display unit 13 is a display screen realized by, for example, a liquid crystal display or an organic EL (Electro-Luminescence) display, and is a display device for displaying various information.
  • the display unit 13 displays information presented from the information processing device 100 under the control of the display control unit 15d.
  • the operation unit 14 functions as an input unit that accepts various operations from the user.
  • the operation unit 14 receives operations on information displayed on the display unit 13 from a user using the terminal device 10.
  • the control unit 15 is realized by the CPU, MPU, or the like executing various programs (for example, the terminal program according to the embodiment) stored in the storage device inside the terminal device 10 using the RAM as a work area. Further, the control unit 15 is realized by, for example, an integrated circuit such as an ASIC or an FPGA.
  • control unit 15 includes an input reception unit 15a, a transmission unit 15b, a response reception unit 15c, and a display control unit 15d, and realizes information processing functions and operations described below. Or run.
  • the internal configuration of the control unit 15 is not limited to the configuration shown in FIG. 5, and may be any other configuration as long as it performs information processing to be described later.
  • the connection relationship between the respective processing units included in the control unit 15 is not limited to the connection relationship shown in FIG. 5, and may be other connection relationships.
  • the input accepting unit 15a accepts information input by the user.
  • the input receiving unit 15a receives information input for inquiring about the service SA.
  • the transmitting unit 15b requests a dialogue starting from the input information by transmitting the input information accepted by the input receiving unit 15a to the information processing device 100.
  • the response reception unit 15c receives information responded from the information processing device 100 in response to information input. For example, in response to determining the question information to be omitted, the response receiving unit 15c receives question information actually presented from the information processing device from among the question information included in the series of question candidates.
  • the display control section 15d causes the display section 13 to display the information accepted by the response reception section 15c.
  • the display control unit 15d controls the display so that information presented from the information processing device 100 is displayed in a manner in which a chatbot responds in accordance with user input information.
  • FIG. 6 is a flowchart showing an accumulation processing procedure for accumulating interaction history.
  • the accumulation process of the dialogue history is performed by the accumulation unit 137.
  • the dialogue history database 124 Prior to explaining the storage method, the data structure of the dialogue history database 124 will be explained.
  • the dialogue history database 124 has items such as "user ID”, “attribute information”, “question ID”, “question information”, “answer ID”, and "answer content”.
  • the "user ID” is identification information that identifies the user who has requested a dialogue by inputting information about the content of the inquiry regarding the service SA.
  • Attribute information is information indicating, for example, gender and age as attributes of the user indicated by the "user ID.”
  • “Question ID” is identification information that identifies the question information presented to the user indicated by “user ID.”
  • “Question information” is a question text presented to the user indicated by “user ID.”
  • the "question information” accumulated as the dialogue history is used to ask the user indicated by the "user ID” whether or not the input information in the form of a question satisfies the condition information corresponding to the input information. It may be limited to the question information used.
  • “Answer ID” is identification information that identifies "answer content.”
  • the “answer content” is information indicating the content of the answer that the user indicated by the “user ID” gave in response to the "question information.”
  • “answer content” may indicate the content of the user's answer as is, or may be an answer category in which the content of the user's answer is categorized.
  • the response categories include a category "yes” indicating affirmation and a category "no" indicating negation.
  • the user (user U1) indicated by the user ID "U1” is a person with the attribute "female/20s", and the user who asked the question identified by the question ID "Q11"
  • An example is shown in which an answer is given to the information (question information Q11) with the answer category "No" identified by the answer ID "ANn”.
  • the storage unit 137 determines whether question information has been presented (step S601). While the question information is not presented (step S601; No), the storage unit 137 waits until the question information is presented.
  • step S601 when the question information is presented (step S601; Yes), the storage unit 137 determines whether the presented question information is question information that is a reply to the input information (question) of the user U4. (Step S602).
  • step S602 If the storage unit 137 determines that the question information is not a question to be asked back (step S602; No), the storage unit 137 moves the process to step S601. On the other hand, if the storage unit 137 determines that the question information is a question to be asked back (step S602; Yes), the storage unit 137 determines whether the user U4 has answered the presented question information (step S603). .
  • step S603 the storage unit 137 waits until the user U4 answers.
  • step S603 if the user U4 answers (step S603; Yes), the storage unit 137 accumulates a dialogue history based on the answer (step S604).
  • the storage unit 137 stores the user ID "U4" that identifies the user U4, the attribute information "female/20s", the question ID "Q11", and the question "Are you hospitalized at a hospital or clinic?" , information in which an answer category corresponding to the answer content of user U4 is associated with an answer ID that identifies this answer category is accumulated as an answer history as shown in FIG. 6.
  • the storage unit 137 may associate the answer category "yes” with the answer ID "ANn".
  • the storage unit 137 may associate the answer category "No” with the answer ID "ANn”. Furthermore, the storage unit 137 can obtain the attribute information of the user U4 by, for example, referring to the user information database 123.
  • FIG. 7 is a flowchart (1) showing the procedure for updating the knowledge graph.
  • the knowledge graph update process is performed between the detection unit 138 and the update unit 139. Further, the update processing explained with reference to FIG. 7 is for more specifically explaining the method of the update processing performed in step S18 of FIG. 3.
  • the detection unit 138 determines whether the dialogue history database 124 has been updated (step S701). As long as the dialogue history database 124 has not been updated (step S701; No), the detection unit 138 waits until it is updated.
  • the detection unit 138 calculates the total number TN1 of users registered in the dialogue history database 124 (i.e., users with a history of dialogue) (Ste S702). Note that the detection unit 138 may determine that the dialogue history database 124 has been updated when a dialogue history is newly accumulated according to the flow described in FIG. 6 .
  • the detection unit 138 extracts the question information Qx newly recorded as the dialogue history (step S703).
  • the following description uses the example of FIG. 6.
  • the question information Q11 is newly recorded as the conversation history because the user U4 who has newly requested a dialogue has answered the question information Q11, so the detection unit 138 detects the newly recorded The question information Q11 is extracted.
  • the detection unit 138 refers to the dialogue history database 124 and calculates the number TN11 of users who have answered the question information Q11 (step S704). Specifically, upon being presented with the question information associated with the question ID "Q11", the detection unit 138 calculates the number TN11 of users who answered the question.
  • the detection unit 138 determines whether the value RA1 indicating the ratio of the number of users TN11 to the total number of users TN1 (TN11/TN1) exceeds a predetermined threshold (step S705). When the detection unit 138 determines that the value RA1 does not exceed the predetermined threshold (step S705; No), the process proceeds to step S701.
  • the detection unit 138 determines that the value RA1 exceeds the predetermined threshold (step S705; Yes), the detection unit 138 assigns these users a specific answer based on the answers of the users who answered the question information Q11.
  • Classify by category step S706.
  • the detection unit 138 classifies these users into an answer category ACy and an answer category ACn based on the content of the answers of the users who answered the question information Q11.
  • the answer category ACy corresponds to the answer content "yes”
  • the answer category ACn corresponds to the answer content "no”. That is, the detection unit 138 classifies users who answered "yes” and users who answered "no" among the users who answered the question information Q11.
  • the detection unit 138 determines whether or not there is an unprocessed answer category ACx among the answer categories used for classification (step S707).
  • the detection unit 138 extracts one of the unprocessed answer categories ACx (step S708). For example, assume that the detection unit 138 has extracted the answer category ACy as the unprocessed answer category ACx. In this case, the detection unit 138 calculates the number TN111 of users who answered the question information Q11 with the answer content belonging to the answer category ACy (ie, "yes") (step S709).
  • the detection unit 138 determines whether the value RA2 indicating the ratio of the number of users TN11 to the number of users TN11 (TN111/TN11) exceeds a predetermined threshold (step S710).
  • the detection unit 138 determines that the value RA2 does not exceed the predetermined threshold (step S710; No)
  • the process proceeds to step S707.
  • the detection unit 138 extracts the unprocessed answer category ACn in step S708, and in step S709, the number of users who answered the question information Q11 with the answer content belonging to the answer category ACn (i.e., "No") TN111 will be calculated.
  • the detection unit 138 determines that the value RA2 exceeds the predetermined threshold (step S710; Yes), the detection unit 138 detects that the question information Q11 tends to be answered with the answer content belonging to the answer category ACx. Detected (step S711).
  • the detection unit 138 extracts the answer category ACy in step S708, and if step S711 has been reached, detects the answer tendency "yes". Specifically, the detection unit 138 detects that the user tends to answer "yes” to the question information Q11.
  • the detection unit 138 extracts the answer category ACn in step S708, and detects the answer tendency "no" if step S711 has been reached. Specifically, the detection unit 138 detects that the user tends to answer "no" to the question information Q11.
  • the updating unit 139 updates the knowledge database 122 based on the response tendency detected in step S711 (step S712).
  • step S707 if the value RA2 does not exceed the threshold for either answer category ACy or answer category ACn, and it is determined that there is no unprocessed answer category ACx (step S707; No), the process is ended.
  • FIG. 8 is a diagram (1) showing an example of a knowledge graph updating method.
  • the update unit 139 updates the knowledge database 122 by linking the answer information AN11 indicating the answer tendency "yes” to the question information Q11 registered as the knowledge graph KG1. Moreover, such a link corresponds to the fact that when the question information Q11 is presented, it is predicted in advance that the answer will be "yes".
  • the update unit 139 updates the knowledge database 122 by linking the answer information AN11 indicating the answer tendency "No" to the question information Q11 registered as the knowledge graph KG1. Moreover, such a link corresponds to the fact that when the question information Q11 is presented, it is predicted in advance that the answer will be "no".
  • FIG. 9 is a flowchart (2) showing the knowledge graph update processing procedure.
  • FIG. 7 describes a method of detecting answer trends and feeding back answer information according to the detection results to the knowledge graph as a predicted answer.
  • FIG. 9 illustrates a method of further detecting attribute trends and feeding back attribute information according to the detection results to the knowledge graph. Therefore, the update processing procedure described in FIG. 9 may be a procedure executed subsequent to step S712 in FIG. 7.
  • the detection unit 138 refers to the dialogue history database 124 and acquires the attribute information of each user who answered the question information Q11 with an answer belonging to the answer category ACy (i.e., "yes") (step S901). .
  • the detection unit 138 classifies the users who answered the question information Q11 with the answer content belonging to the answer category ACy into each predetermined attribute category (step S902). For example, the detection unit 138 may classify these users into “20s”, “30s”, “40s”, . . . , “male”, “female”, etc. based on the attribute information.
  • the detection unit 138 classifies users who answered the question information Q11 with answer contents belonging to the answer category ACy based on the attribute information into the attribute category DCa and the attribute category DCs. shall be classified.
  • the attribute category DCa corresponds to "20s”
  • the attribute category DCs corresponds to "female.”
  • the detection unit 138 classifies users who have answered "yes” to question information Q11 into users who belong to "20s” and users who belong to "female.”
  • the detection unit 138 detects that among the users who answered "No” to the question information Q11, users who belong to “20s” and “Female” ”, and further processing will be performed.
  • the detection unit 138 determines whether or not there is an unprocessed attribute category DCx among the attribute categories used for classification (step S903).
  • the detection unit 138 extracts one of the unprocessed attribute categories DCx (step S904). For example, assume that the detection unit 138 has extracted the attribute category DCa as the unprocessed attribute category DCx. In this case, the detection unit 138 calculates the number TN1111 of users classified by the attribute category DCa (users belonging to the attribute category DCa) in step S902 (step S905).
  • the detection unit 138 determines that a value RA3 indicating the ratio (TN1111/TN111) of the number of users TN111 to the number TN111 of users who answered with answer content belonging to the answer category ACy with respect to the question information Q11 exceeds a predetermined threshold value. It is determined whether or not (step S906). If the detection unit 138 determines that the value RA3 does not exceed the predetermined threshold (step S906; No), the process proceeds to step S903. In this case, the detection unit 138 extracts the unprocessed attribute category DCs in step S904, and calculates the number TN1111 of users classified by the attribute category DCs (users belonging to the attribute category DCs) in step S905.
  • step S906 determines that the value RA3 exceeds the predetermined threshold (step S906; Yes)
  • the detection unit 138 extracts the attribute category DCa in step S903, and if the process has reached step S907, detects the attribute tendency "20s". Specifically, the detection unit 138 detects that users who answer "yes" to question information Q11 tend to be in their "20s.”
  • the detection unit 138 extracts the attribute category DCs in step S903, and if the process has reached step S907, the detection unit 138 determines that users who answer "no" to the question information Q11 tend to be "female". To detect.
  • the updating unit 139 updates the knowledge database 122 based on the attribute tendency detected in step S907 (step S908).
  • step S903 if the value RA3 does not exceed the threshold in either attribute category DCa or attribute category DCs, and it is determined that there is no unprocessed attribute category DCx (step S903; No), the process is ended.
  • FIG. 10 is a diagram (2) illustrating an example of a knowledge graph updating method.
  • the update unit 139 updates the knowledge database 122 by linking the attribute information DT11 indicating the attribute tendency "20s" to the question information Q11 registered as the knowledge graph KG1.
  • answer information AN11 indicating an answer tendency of "yes” has already been linked to question information Q11. If an attribute tendency is further detected in such a state, the updating unit 139 further updates the knowledge database 122 by also linking attribute information indicating the attribute tendency to the question information Q11.
  • the update unit 139 links the attribute information DT11 indicating the attribute tendency "20s” to the question information Q11 registered as the knowledge graph KG1.
  • the attribute tendency "female” is detected in step S908.
  • the updating unit 139 links attribute information DT11 indicating the attribute tendency "female” to the question information Q11.
  • the detection unit 138 may not be able to detect the attribute tendency because the value RA3 does not exceed the threshold value in either the attribute category DCa or the attribute category DCs.
  • the updating unit 139 may link non-attribute information DT12 indicating all attributes without being limited to a specific attribute category to the question information Q11, as shown in FIG. 10(b).
  • FIG. 11 is a flowchart showing a determination processing procedure for determining question information to be omitted.
  • a scene is shown in which the user U5 interacts with the chatbot CB by utilizing the dialog support system DL in order to enjoy the service SA.
  • the user U5 uses the terminal device 10 to input the following information: "I was hospitalized due to a traffic accident. Will I receive benefits?" and sends the message to the dialogue support system DL. I am inquiring.
  • the terminal device 10 transmits input information IN11 in the form of a question input by the user U5 to the information processing device 100.
  • the analysis unit 132 estimates the slot by performing syntax analysis and semantic analysis on the input information IN11 (step S1101). For example, the analysis unit 132 can estimate "traffic accident", “hospitalization”, etc. as slots.
  • the extraction unit 133 refers to the knowledge database 122 and extracts a series of question candidates that correspond to the slot (step S1102). For example, the extraction unit 133 refers to the knowledge graph KG1 and identifies a plurality of pieces of condition information depending on the slot. Then, the extraction unit 133 extracts, for each piece of specified condition information, question information that asks whether or not the condition information is satisfied. Then, the extraction unit 133 defines the extracted question information group as a series of question candidates that can be presented to the user U5.
  • the extraction unit 133 extracts question information Q11 with the content “Are you hospitalized at a hospital or clinic?” and question information Q12 with the content “Are you hospitalized for 5 days or more?”
  • question information Q11 and Q12 can be information for asking questions in response to the content of the question indicated by the input information IN11.
  • the filter processing unit 134 determines whether unprocessed question information Qx exists among the question information Q11 and Q12 (step S1103).
  • the filter processing unit 134 extracts one of the unprocessed question information Qx from the knowledge graph KG1 (step S1104).
  • the filter processing unit 134 determines whether meta information indicating a predicted answer is attached to the question information Q11 (step S1105). Specifically, the filter processing unit 134 determines whether answer information indicating an answer tendency is added as meta information to the question information Q11.
  • step S1105 If meta information has not been added (step S1105; No), the filter processing unit 134 moves the process to step S1103.
  • the knowledge graph KG1 in the state shown in FIG. It will be determined that it is.
  • the filter processing unit 134 determines that meta information indicating a predicted answer is attached (step S1105; Yes), the filter processing unit 134 in turn applies meta information indicating an attribute tendency to the question information Q11. It is determined whether the information is provided as information (step S1106).
  • the filter processing unit 134 determines that the attribute information is not provided as meta information (step S1106; No), and sets the question information Q11 to the question to be omitted. It is determined as information (step S1108), and the process moves to step S1103. Note that this flow of processing allows the question information Q11 to be self-evident to the user U5 even if the attribute tendency is not detected, as long as it is possible to predict the answer of the user U5 by detecting the answer tendency. This is based on the idea that it can be skipped because it contains question information.
  • the filter processing unit 134 determines that the attribute information (attribute information DT11) is added as meta information. It turns out.
  • the filter processing unit 134 determines whether attribute information is provided (step S1106; Yes), the filter processing unit 134 determines whether the user U5 belongs to the attribute category indicated as the attribute tendency in this attribute information. (Step S1107). For example, the filter processing unit 134 may determine that the user belongs to the attribute category if similarity is obtained by comparing the attribute category and the user's attribute information.
  • the attribute information DT11 indicates the attribute tendency "20s". Therefore, if the user U5 is a person in his or her 20s, the filter processing unit 134 determines that the user U5 belongs to the attribute category (step S1107; Yes), and determines the question information Q11 as the question information to be omitted. do. On the other hand, if the user U5 is a person in his or her 30s, the filter processing unit 134 determines that the user U5 does not belong to the attribute category (step S1107; No) and does not set the question information Q11 to be omitted. , the process moves to step S1103.
  • step S1104 the case where the question information Q11 is extracted as the unprocessed question information Qx in step S1104 has been described as an example.
  • step S1103 for example, if question information Q12 is extracted as unprocessed question information Qx, processing is performed on question information Q12 from step S1105 onwards.
  • the filter processing unit 134 determines that there is no unprocessed question information Qx (step S1103; No)
  • the filter processing unit 134 selects the question to be omitted from among the question information Q11 and Q12 included in the series of question candidates.
  • Question information other than this information is determined as question information to be presented (step S1109).
  • the filter processing unit 134 determines both question information Q11 and Q12 as question information to be presented. Such a pattern will be referred to as a "first pattern.”
  • the filter processing unit 134 determines only the question information Q12 as the question information to be presented. Such a pattern will be referred to as a "second pattern.”
  • the generation unit 135 generates question information TQx to be actually presented based on the question information to be presented (step S1110). For example, in the first pattern, the generation unit 135 generates question information TQ11 as the question information TQx to be actually presented based on the question information Q11, and generates question information TQ1 as the question information TQx to be actually presented based on the question information Q12. Question information TQ12 is generated.
  • the generation unit 135 only generates question information TQ12 based on question information Q12.
  • the output control unit 136 performs output control so that the question information TQx is presented to the user U5 (step S1111). For example, the output control unit 136 may transmit the question information TQx to the terminal device 10 of the user U5 so that the question information TQx is displayed on the display unit 13 included in the terminal device 10.
  • Step S1112 the storage unit 137 executes an accumulation process of accumulating the question information TQx and the answer as a dialogue history.
  • the method of accumulating the dialogue history has already been explained with reference to FIG. 6, so the explanation will be omitted.
  • the dialogue history database 124 is updated due to the accumulation of dialogue history, the knowledge graph is also updated according to the update result, but this method has been explained with reference to FIGS. 7 to 9 and will therefore be omitted.
  • FIG. 12 is a diagram showing an example of a display screen when question information is displayed.
  • FIG. 12 an example of a display screen will be described using the example of FIG. 11.
  • FIG. 12A shows an example of a display screen corresponding to the "first pattern" shown in step S1109 of FIG. 11.
  • FIG. 12(b) shows an example of a display screen corresponding to the "second pattern" shown in step S1109 of FIG. 11.
  • question information TQ11 generated from question information Q11 is displayed on the display screen of the terminal device 10. Specifically, question information TQ11 is displayed in such a manner that the chatbot CB asks back the input information IN11 of the user U5. As shown in FIG. 12(a), the question information TQ11 includes question information Q11 with the content "Are you hospitalized at a hospital or clinic?"
  • the question information TQ11 may be provided with a button for answering "yes” to the question information Q11 and a button for answering "no". Therefore, for example, when the user U5 answers using any button, the question information TQ12 generated from the question information Q12 is displayed on the terminal device 10 as shown in FIG. 12(a). displayed on the screen. As shown in FIG. 12(a), the question information TQ12 includes question information Q11 with the content "Are you hospitalized for 5 days or more?"
  • the question information TQ12 may be provided with a button for answering "yes” to the question information Q12 and a button for answering "no". Therefore, when the user U5 answers using any button, the answer comprehensively determined from the answer to the question information Q11 and the answer to the question information Q12 is the official answer to the input information IN11. May be displayed as an answer.
  • question information TQ12 generated from question information Q12 is displayed on the display screen of the terminal device 10, as shown in FIG. 12(b). That is, compared to FIG. 12(a), the display of question information TQ11 (question information Q11) is skipped and only question information TQ12 is displayed. Specifically, question information TQ12 is displayed in such a manner that the chatbot CB answers the input information IN11 of the user U5. As shown in FIG. 12(b), the question information TQ12 includes question information Q11 with the content "Are you hospitalized for 5 days or more?"
  • the question information TQ12 may be provided with a button for answering "yes” to the question information Q12 and a button for answering "no". Therefore, when the user U5 answers using any button, the answer comprehensively determined from the answer to the question information Q11 and the answer to the question information Q12 is the official answer to the input information IN11. May be displayed as an answer.
  • the content of the answer to the question information Q11 is predicted based on the answer information AN11 given to the question information Q11 in the knowledge graph KG1.
  • the output control unit 136 may control the question information to be omitted to be presented to the user if a predetermined condition is satisfied.
  • FIG. 13 is a diagram illustrating an example of a situation in which question information to be omitted is presented.
  • FIG. 13 shows the example of FIG. 12(b) in which the presentation of question information Q11 among question information Q11 and Q12 is skipped and only question information Q12 is presented to user U5.
  • the output control unit 136 sends a notification indicating that there is skipped information and is controlled to request question information to be omitted (in this example, question information Q11).
  • NT is further displayed.
  • link information L for accessing the question information to be omitted may be attached to the notification NT.
  • the output control unit 136 may display the question information Q11.
  • the output control unit 136 may display question information TQ11 shown in FIG. 12(a).
  • the storage process described above may also be performed using the content of the answer.
  • the output control unit 136 notifies the content for which skipped question information can be requested, and when the skipped question information is requested using such content, this question information is presented anew.
  • the output control unit 136 may control the question information to be omitted to be presented to the user if the timing at which the user answers is a predetermined timing based on the tendency obtained from the interaction history. . This point will be explained using the example of FIG. 9, assuming that the user U5 is a person in his 20s.
  • the output control unit 136 may perform a lottery and present the question information Q11 with a probability of 2/10. For example, the output control unit 136 performs a lottery based on a user U5 who is in his 20s requesting a dialogue, and if the probability of 2/10 is won, the output control unit 136 presents the question information Q11 without skipping. good.
  • the knowledge graph KG1 in the knowledge database 122 is updated as needed according to the accumulation of dialogue history.
  • the updating unit 139 may clear the update status of the knowledge graph KG1 and return it to the initial state, for example, if a predetermined condition is satisfied.
  • the update unit 139 may clear the update status of the knowledge graph KG1 and return it to the initial state at every specific cycle.
  • the update unit 139 may clear the update status of the knowledge graph KG1 every six months, or may clear the update status every season.
  • the updating unit 139 updates the knowledge graph K1 with the trend information indicating the tendency detected from the external data.
  • answer information AN11 is linked to question information Q11 of knowledge graph K1
  • the answer tendency is "yes".
  • more formal content for example, a face-to-face dialogue with a service SA representative, formal documents
  • the updating unit 139 may preferentially link the answer information AN11 indicating the answer tendency "No” instead of the answer tendency "Yes” to the question information Q11.
  • FIG. 14 is a block diagram illustrating an example of a hardware configuration of a computer corresponding to an apparatus according to an embodiment and a modification of the present disclosure. Note that FIG. 14 shows an example of a hardware configuration of a computer corresponding to the apparatus according to the embodiment and modification of the present disclosure, and the configuration is not limited to that shown in FIG. 14.
  • the computer 1000 includes a CPU (Central Processing Unit) 1100, a RAM (Random Access Memory) 1200, a ROM (Read Only Memory) 1300, an HDD (Hard Disk Drive) 1400, a communication interface 1500, and an input/output It has an interface 1600.
  • CPU Central Processing Unit
  • RAM Random Access Memory
  • ROM Read Only Memory
  • HDD Hard Disk Drive
  • the CPU 1100 operates based on a program stored in the ROM 1300 or the HDD 1400 and controls each part. For example, the CPU 1100 loads programs stored in the ROM 1300 or HDD 1400 into the RAM 1200, and executes processes corresponding to various programs.
  • the ROM 1300 stores boot programs such as BIOS (Basic Input Output System) that are executed by the CPU 1100 when the computer 1000 is started, programs that depend on the hardware of the computer 1000, and the like.
  • BIOS Basic Input Output System
  • the HDD 1400 is a computer-readable recording medium that non-temporarily records programs executed by the CPU 1100 and data used by the programs. Specifically, HDD 1400 records program data 1450.
  • the program data 1450 is an example of an information processing program for realizing the information processing method according to the embodiment and modification of the present disclosure, and data used by the information processing program.
  • Communication interface 1500 is an interface for connecting computer 1000 to external network 1550 (eg, the Internet).
  • CPU 1100 receives data from other devices or transmits data generated by CPU 1100 to other devices via communication interface 1500.
  • the input/output interface 1600 is an interface for connecting the input/output device 1650 and the computer 1000.
  • CPU 1100 receives data from an input device such as a keyboard or mouse via input/output interface 1600. Further, the CPU 1100 transmits data to an output device such as a display device, a speaker, or a printer via the input/output interface 1600.
  • the input/output interface 1600 may function as a media interface that reads programs and the like recorded on a predetermined recording medium. Examples of media include optical recording media such as DVD (Digital Versatile Disc) and PD (Phase change rewritable disk), magneto-optical recording media such as MO (Magneto-Optical disk), tape media, magnetic recording media, semiconductor memory, etc. It is.
  • the CPU 1100 of the computer 1000 executes the information processing program loaded on the RAM 1200. Accordingly, various processing functions executed by each part of the control unit 130 shown in FIG. 4 are realized. That is, the CPU 1100, the RAM 1200, etc. cooperate with software (information processing program loaded on the RAM 1200) to perform an information processing method by the apparatus (for example, the information processing apparatus 100) according to the embodiments and modifications of the present disclosure. Realize.
  • the CPU 1100 of the computer 1000 executes the terminal program loaded on the RAM 1200.
  • Various processing functions executed by each part of the control unit 15 shown in FIG. 5 are realized. That is, the CPU 1100, the RAM 1200, etc., cooperate with software (information processing program loaded on the RAM 1200) to execute the information processing method by the device (as an example, the terminal device 10) according to the embodiment and modification of the present disclosure. Realize.
  • an acquisition unit that acquires a series of question candidates that can be presented to a requester who has requested a dialogue from among the question information registered in the knowledge database; A question to be omitted from which presentation to the requester is omitted from among the question information included in the series of question candidates, based on the history of interactions with the user realized by providing the question information to the user.
  • An information processing device comprising: a determining unit that determines information; (2) The knowledge database includes, as the question information, graph-structured question information generated based on the terms and conditions of a predetermined service, which is used for dialogue with a user who wants to enjoy the predetermined service.
  • the acquisition unit acquires a series of question candidates that can be presented to a requester who has input a question regarding the predetermined service, from among the question information registered in the knowledge database.
  • Processing equipment (3)
  • the question information includes condition information for determining an answer to the question input by the requester,
  • the acquisition unit acquires, as the series of question candidates, a series of question sentences for asking whether or not the condition information corresponding to the question is satisfied, from among the condition information;
  • the determining unit determines a question sentence to be omitted from being presented to the requester from among the series of question sentences based on a user's interaction history with respect to the series of question sentences.
  • the determining unit determines that, based on the tendency obtained from the dialogue history, there is question information in the series of question candidates that allows predicting the response content of the requester. If so, the information processing device according to (1) above, determines the question information as the question information to be omitted.
  • a detection unit that detects a tendency regarding the content of answers given by the user to the question information based on the dialogue history;
  • the information processing device according to (4) further comprising: an updating unit that updates the knowledge database based on the trend.
  • the detection unit detects an answer tendency of what kind of answers the user tends to give to the question information based on the interaction history, The information processing device according to (5), wherein the updating unit updates the knowledge database based on the response tendency.
  • the detection unit classifies users who answered the question information by answer content, and detects response trends based on the number of users for each answer content, When it is detected that the user tends to answer any of the answer contents used for classification, the updating unit links answer information indicating the detected answer tendency to the question information.
  • the information processing device according to (6) above, wherein the information processing device is registered in a knowledge database in a state.
  • the determining unit is configured to assign the answer information to the question information.
  • the information processing device according to (7), wherein the requester predicts that the requester will answer with the answer content indicated by and determines the question information as the question information to be omitted.
  • the detection unit detects an attribute tendency of a user who answered with the answer content indicated by the answer tendency based on the interaction history, The information processing device according to (7), wherein the updating unit updates the knowledge database based on the attribute tendency.
  • the detection unit classifies the users who answered with the answer content according to predetermined attributes, and detects an attribute tendency based on the number of users for each attribute, If it is detected that a user who has any of the attributes used for classification tends to answer with the answer content, the updating unit updates attribute information indicating the detected attribute tendency.
  • the information processing device registers the information in a knowledge database in a state where the information is linked to the question information.
  • the determining unit determines the linked attribute information; The information processing device according to (10), wherein it is determined whether or not the question information is to be omitted based on similarity with attribute information possessed by the requester. (12) If the determination unit determines that there is similarity between the attribute information linked to the question information and the attribute information possessed by the requester, the determination unit sets the question information as question information to be omitted.
  • the information processing device according to (11) above.
  • the detection unit further detects a trend regarding the user's answer content from predetermined external data regarding the question information, When the tendency indicated by the information linked to the question information and the tendency detected from the external data diverge, the updating unit updates the trend information indicating the tendency detected from the external data to the question information.
  • the information processing device according to (7) or (10) above.
  • the apparatus Even if the question information to be omitted is determined, if a predetermined condition is satisfied, the apparatus further includes an output control unit configured to control the question information to be omitted to be presented to the requester.
  • the information processing device according to (1).
  • the output control unit sends a notification that there is information whose presentation has been omitted, and requests the question information to be omitted.
  • (14) above in which a controllable notification is presented to the requester, and when the question information to be omitted is requested, the question information to be omitted is controlled to be presented to the requester; The information processing device described.
  • the output control unit controls the question information to be omitted to be presented to the requester when the requester answers at a predetermined timing based on a tendency obtained from the dialogue history.
  • the information processing device according to (14) above.
  • An information processing method executed by an information processing device comprising: an acquisition step of acquiring a series of question candidates that can be presented to a requester who has requested a dialogue from among the question information registered in the knowledge database; A question to be omitted from which presentation to the requester is omitted from among the question information included in the series of question candidates, based on the history of interactions with the user realized by providing the question information to the user.
  • An information processing method comprising: a decision step of determining information; (18) an acquisition unit that acquires a series of question candidates that can be presented to a requester who has requested a dialogue from among the question information registered in the knowledge database; A question to be omitted from which presentation to the requester is omitted from among the question information included in the series of question candidates, based on the history of interactions with the user realized by providing the question information to the user.
  • a terminal device that communicates with an information processing device comprising: a determining unit that determines information, and is used by the requester; a reception unit that receives question information actually presented from the information processing device from among the question information included in the series of question candidates in response to the determination of the question information to be omitted;
  • a terminal device comprising: a display control unit that displays presented question information on a screen. (19) an acquisition unit that acquires a series of question candidates that can be presented to a requester who has requested a dialogue from among the question information registered in the knowledge database; A question to be omitted from which presentation to the requester is omitted from among the question information included in the series of question candidates, based on the history of interactions with the user realized by providing the question information to the user.
  • a terminal program executed by a terminal device used by the requester and communicating with an information processing device comprising: A reception procedure for receiving question information actually presented from the information processing device from among the question information included in the series of question candidates in response to the determination of the question information to be omitted; A terminal program for causing the terminal device to execute a display control procedure for displaying the presented question information on a screen.
  • Information processing system 10 Terminal device 11 Communication unit 12 Storage unit 13 Display unit 14 Operation unit 15 Control unit 15a Input reception unit 15b Transmission unit 15c Response reception unit 15d Display control unit 100 Information processing device 110 Communication unit 120 Storage unit 121 Terms and conditions information Database 122 Knowledge database 123 User information database 124 Dialogue history database 130 Control unit 131 Reception unit 132 Analysis unit 133 Extraction unit 134 Filter processing unit 135 Generation unit 136 Output control unit 137 Accumulation unit 138 Detection unit 139 Update unit

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Technology Law (AREA)
  • Computational Linguistics (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

Un dispositif de traitement d'informations (100) selon un mode de réalisation comprend : une unité d'acquisition (133) qui acquiert une série de questions candidates qui peuvent être présentées à un demandeur qui a demandé une interaction, parmi des informations de question enregistrées dans une base de données de connaissances ; et une unité de détermination (134) qui détermine, sur la base d'un historique d'interaction avec un utilisateur obtenu en fournissant des informations de question à l'utilisateur, des informations de question à omettre de la présentation au demandeur parmi les informations de question incluses dans la série de questions candidates.
PCT/JP2023/025063 2022-07-13 2023-07-06 Dispositif de traitement d'informations, procédé de traitement d'informations, dispositif terminal et programme de terminal WO2024014383A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2022-112378 2022-07-13
JP2022112378 2022-07-13

Publications (1)

Publication Number Publication Date
WO2024014383A1 true WO2024014383A1 (fr) 2024-01-18

Family

ID=89536631

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2023/025063 WO2024014383A1 (fr) 2022-07-13 2023-07-06 Dispositif de traitement d'informations, procédé de traitement d'informations, dispositif terminal et programme de terminal

Country Status (1)

Country Link
WO (1) WO2024014383A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019187463A1 (fr) * 2018-03-27 2019-10-03 株式会社Ｎｔｔドコモ Serveur de dialogue
JP2019207648A (ja) * 2018-05-30 2019-12-05 株式会社野村総合研究所 対話型業務支援システム
JP2020080025A (ja) * 2018-11-13 2020-05-28 株式会社日立製作所 質問応答データ生成装置および質問応答データ生成方法
JP2020184294A (ja) * 2019-04-26 2020-11-12 Ａｒｉｔｈｍｅｒ株式会社 対話管理サーバ、対話管理方法、及びプログラム
US20210081442A1 (en) * 2019-09-12 2021-03-18 Intuit Inc. System and method for reducing user query ambiguity through chatbot clarifying questions

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019187463A1 (fr) * 2018-03-27 2019-10-03 株式会社Ｎｔｔドコモ Serveur de dialogue
JP2019207648A (ja) * 2018-05-30 2019-12-05 株式会社野村総合研究所 対話型業務支援システム
JP2020080025A (ja) * 2018-11-13 2020-05-28 株式会社日立製作所 質問応答データ生成装置および質問応答データ生成方法
JP2020184294A (ja) * 2019-04-26 2020-11-12 Ａｒｉｔｈｍｅｒ株式会社 対話管理サーバ、対話管理方法、及びプログラム
US20210081442A1 (en) * 2019-09-12 2021-03-18 Intuit Inc. System and method for reducing user query ambiguity through chatbot clarifying questions

Similar Documents

Publication Publication Date Title
JP7209818B2 (ja) 自動ナビゲーションを容易にするためのウェブページの分析
US20190272269A1 (en) Method and system of classification in a natural language user interface
US11113672B2 (en) Computer support for meetings
US10521189B1 (en) Voice assistant with user data context
EP3511887A1 (fr) Assistant automatisé de conversation pour la fourniture de données interactives à l'aide d'un traitement du langage naturel (npl) - système et procédé
EP4213043A1 (fr) Fourniture de suggestions de paquets de commandes pour un assistant automatisé
US20210303607A1 (en) System to assist users of a software application
CN108153800B (zh) 信息处理方法、信息处理装置以及记录介质
US11580112B2 (en) Systems and methods for automatically determining utterances, entities, and intents based on natural language inputs
US8744903B2 (en) Event resolution
US11482223B2 (en) Systems and methods for automatically determining utterances, entities, and intents based on natural language inputs
US20210406913A1 (en) Metric-Driven User Clustering for Online Recommendations
US11632341B2 (en) Enabling communication with uniquely identifiable objects
EP1500015A2 (fr) Rapport d'interactions avec des clients
US9614790B2 (en) Apparatus for controlling execution of software, method for controlling thereof, and computer-readable recording medium having computer program for controlling thereof
US20090150341A1 (en) Generation of alternative phrasings for short descriptions
US9946971B1 (en) Method for the management of artifacts in knowledge ecosystems
US10929485B1 (en) Bot search and dispatch engine
JP6899797B2 (ja) 問合せ機器特定システム、問合せ機器特定方法
US20230409836A1 (en) Query construction from conversational inputs
US20210295409A1 (en) Unified alternative item selection user experience
WO2024014383A1 (fr) Dispositif de traitement d'informations, procédé de traitement d'informations, dispositif terminal et programme de terminal
US11206236B2 (en) Systems and methods to prioritize chat rooms using machine learning
CN112395517A (zh) 房源搜索展示方法、装置及计算机可读存储介质
CN112352223A (zh) 用于输入建议的方法和系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23839555

Country of ref document: EP

Kind code of ref document: A1