Information processing device, information processing method, and program

Abstract

We learn the thought processes of users who make final decisions regarding metrics and generate multiple personas. [Solution] The information processing device is configured to perform the following actions: query the case data DB based on initial dialogue information received from the user terminal to read case data associated with the initial dialogue information, generate a first group of personas based on the case data, obtain user evaluations from the user terminal for each combination of elements and element values ​​of each persona in the first group, update the read case data based on the determination of whether each combination of elements and element values ​​of each persona in the first group is a positive or negative user evaluation based on the user evaluation, and generate a second group of personas that are tailored to the individual circumstances of the user based on the updated case data.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program. More specifically, the present invention relates to an information processing apparatus, an information processing method, and a program that learn the thinking of a user who determines a final index and generate a multiplexed persona personalized thereby.

Background Art

[0002] There is known a technique for generating a "persona" by imitating a human (Patent Document 1). A persona may be called an interactive AI (artificial intelligence), and is a technique in which a computer can perform recognition, inference, judgment, prediction, proposal, etc., similar to human intellectual activities, based on a rule tree (decision tree). In this technique, personas generated for each aspect (work, family, region, individual, hobby, etc.) constituting a human can be treated as an aggregate, and each persona can include one or more rule trees. In the system of Patent Document 1, it is shown that a persona of a human to be imitated, called a donor, is generated, and a general user can access the system using his / her own device and have a conversation with the persona.

[0003] The system of Patent Document 1 represents an example based on input data as a set of an element (conditional attribute) and a target element (conclusion attribute), and also learns a rule for estimating a target element value (conclusion attribute value) from an element value (conditional attribute value) of an element observed based on the example to generate a rule tree. The system of Patent Document 1 outputs a target element value by automatically traversing the rule tree based on the way of response (element value) of a preset question item (element).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

[0005] In the system described in Patent Document 1, the personification system generates personas of experts in various fields and provides them to general users as conversation partners, thereby enabling general users to receive consulting. However, the various problems in modern society are complex. While problems that people face may sometimes appear to be classifiable into a specific theme, in reality, various elements are mixed together, and problems arise from the overlapping of various causes. When people converse about a certain theme or problem, the topic may shift to another theme or problem when a specific keyword appears. In such cases, consulting cannot be conducted solely by experts in a specific field; the cooperation of experts in other fields must be obtained.

[0006] One might consider creating a digital chimera by combining the personas of multiple experts to address various problems. However, simply combining the personas of multiple experts will not adequately address the problems faced by general users. For example, in the medical field, even when prescribing the same drug, the criteria for prescribing may differ among doctors, and the rules for administering prescribed drugs may differ among nurses. Furthermore, even when prescribing and administering the same drug for the same disease, the rules may differ from one medical institution to another.

[0007] Therefore, simply combining the personas of multiple experts results in a massive rule tree. While a massive rule tree has the advantage of being able to handle multiple ways of thinking, (1) Because it requires many examples to generate, it takes an enormous amount of time to complete. (2) Considering the possibility of branching out into various themes and issues, the amount of information (elements) required for a single case becomes excessive, and, (3) If it is discovered after the fact that necessary elements are missing, or if additional elements are found after the fact, it will be necessary to add those elements to all cases. These are some of the disadvantages. For this reason, simply combining the personas of multiple experts is likely to create significant operational challenges. In the example of the medical field mentioned above, the selection and administration of prescribed drugs may differ depending on individual circumstances such as the health condition of the patient (the general user) and their relationships with family and others around them.

[0008] Therefore, there is a need to create tailor-made digital chimeras (a collection of multiple personas) by combining the personas of multiple experts based on the individual circumstances of the average user.

[0009] This invention was made to solve these problems, and aims to provide an information processing device, an information processing method, and a program that learn the thinking of users who make final decisions on indicators and generate multiple personas based on those personas. [Means for solving the problem]

[0010] To solve the above problems, the information processing apparatus according to the present invention is: An information processing device comprising a control unit and a storage unit, The memory unit includes a case data DB that stores elements and case information associated with experts, which serve as the basis for generating personas. The control unit, Based on the initial dialogue information received from the user terminal, the system queries the case data database to retrieve the case data associated with the initial dialogue information. Based on the aforementioned case data that has been read out, a first expert persona is generated, To generate a first group of personas, including the first expert persona, Obtain user evaluations from the user terminal for each combination of elements and element values ​​of each persona in the first persona group, Based on the user evaluations, determine whether each combination of elements and element values ​​for each persona in the first persona group is positive or negative in terms of user evaluations. Based on the above determination, the retrieved case data is updated, To generate a first revised expert persona based on the updated case data, To generate user personas based on combinations of elements and element values ​​that result in positive user evaluations, and combinations of elements and element values ​​that result in negative user evaluations, To generate a second group of personas, including the first modified expert persona and the user persona, It is configured to execute. [Effects of the Invention]

[0011] According to the present invention, it becomes possible to adjust individual personas included in a group of personas associated with the service content desired by the user to suit the user's individual circumstances. Furthermore, when conducting a dialogue with a user, it becomes possible to estimate whether a positive or negative evaluation is likely to be obtained if the dialogue proceeds in a certain direction. In addition, when generating a group of personas to be used in a dialogue to prevent the dialogue from resulting in a negative evaluation, it becomes possible to generate the desired personas by selecting and discarding case data. [Brief explanation of the drawing]

[0012] A detailed understanding of the embodiments disclosed herein can be obtained from the following description illustrated in relation to the accompanying drawings. [Figure 1] This diagram illustrates the overall system configuration including the information processing device 10, user terminal 11, and expert terminal 12 according to the present invention. [Figure 2] This is a system configuration diagram of the information processing device 10 according to the present invention. [Figure 3]FIG. is a diagram showing an example of elements of each expert and case information stored in the case data information 106, and an example of a rule tree of a persona of each expert generated therefrom. [Figure 4] FIG. is a diagram for explaining a rule tree. [Figure 5] FIG. is a flowchart for explaining a process in which a first multiplexed persona is generated through an initial interaction between the user terminal 11 and the information processing apparatus 10, and then a second multiplexed persona adapted to the individual circumstances of the user is generated through an expert interaction. [Figure 6] FIG. is a diagram showing an example of merged case data and an example of an expert persona generated from the merged case data. [Figure 7] FIG. is a diagram showing an example of a first multiplexed persona group and a second multiplexed persona group. [Figure 8] FIG. is a diagram for explaining a process of adjusting a combination of elements and element values of each persona included in the second persona group. [Figure 9] FIG. is a diagram exemplifying whether a combination of each target element and target element value is a positive evaluation or a negative evaluation. [Figure 10] FIG. is a diagram showing an example of a first modified expert persona and a second modified expert persona.

MODE FOR CARRYING OUT THE INVENTION

[0013] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The same reference numerals in the plurality of drawings represent the same elements, and redundant descriptions will be omitted. The examples given in this specification are merely examples and are not intended to impose any limitations by the description of this specification.

[0014] In this specification, services provided by professionals include medical services, legal services, and accounting and tax services. Medical services can be provided by a team of professionals such as doctors, dentists, dental hygienists, nurses, public health nurses, physical therapists, occupational therapists, radiologists, and pharmacists. Legal services can be provided by a team of professionals such as lawyers, judicial scriveners, patent attorneys, and administrative scriveners. Accounting and tax services can be provided by a team of professionals such as certified public accountants, tax accountants, social insurance labor consultants, and small and medium-sized enterprise consultants. In this specification, medical services are used as an example to illustrate the invention, but the present invention can be applied to various professional services.

[0015] (Overall structure and the function of each component) Figure 1 illustrates the overall system configuration including the information processing device 10, user terminal 11, and expert terminal 12 according to the present invention. The information processing device 10 is connected to the user terminal 11 and expert terminal 12 so as to be able to communicate with each other via any network. In Figure 1, only one user terminal 11 and one expert terminal 12 are shown, but there may be multiple such terminals.

[0016] The information processing device 10 provides a dialogue application to the user terminal 11 for interaction between the user and the information processing device 10. The information processing device 10 can communicate text information, voice information, and / or image information with the user terminal 11 via the dialogue application. In response to access from the user terminal 11, the information processing device 10 provides the user terminal 11 with the initial dialogue screen (first screen) of the dialogue application. The initial dialogue screen provides an interface in which the user can input or select what they want to discuss. For example, in the case of medical services, the user can input or select information such as the reason for visiting, symptoms (chief complaint), severity of the reason for visiting, specific location of symptoms, progression of symptoms, medical history, and desired medical department. The input or selected information is called initial dialogue information.

[0017] When the information processing device 10 receives initial dialogue information from the user terminal 11 via the initial dialogue screen, it analyzes the initial dialogue information. Based on the analysis results of the initial dialogue information, the information processing device 10 reads case data associated with the initial dialogue information and generates a first expert persona based on each of the case data. The information processing device 10 can merge the read case data and generate a second expert persona based on the merged case data. The first expert persona plays the role of an individual expert, while the second expert persona plays the role of an expert team. The information processing device 10 can combine the first expert persona and / or the second expert persona into a first multiplexed persona group.

[0018] The information processing device 10 receives and stores case data associated with each expert from the expert terminal 12. Based on the stored case data, the information processing device 10 can generate a first expert persona, which is the persona of each expert, and stores the data of the first expert persona (rule tree). Based on predetermined conditions, the information processing device 10 can merge the stored case data and generate a second expert persona, which is the team of experts, based on the merged case data.

[0019] The information processing device 10 can enable the user to interact with the first group of multiple personas through the expert dialogue screen (second screen) of the dialogue application. The information processing device 10 presents the user terminal 11 with inquiries about each combination of elements and element values ​​included in each of the first group of multiple personas, and obtains responses from the user. The user's responses may be positive user evaluations (e.g., "OK") or negative user evaluations (e.g., "NG"). More specifically, during the conversation through the expert dialogue screen, the information processing device 10 repeatedly traverses the elements and element values ​​of the rule trees of the first and second expert personas based on the above inquiries and responses, and can determine which combinations of elements and element values ​​result in positive or negative user evaluations. Based on the positive and negative user evaluations for each combination of elements and element values, the information processing device 10 can generate user personas.

[0020] The information processing device 10 updates the case data based on combinations of elements and element values ​​that result in positive user evaluations, and combinations of elements and element values ​​that result in negative user evaluations. The information processing device 10 generates a first correction expert persona based on the updated case data. The information processing device 10 merges the updated case data and generates a second correction expert persona based on the merged updated case data. The information processing device 10 can combine the first correction expert persona, the second correction expert persona, and / or user personas into a second multiplexed persona group.

[0021] The user terminal 11 may be any type of device capable of operating in a wired or wireless environment (e.g., a smartphone, PC, tablet, etc.) and is not limited to any specific device. The user terminal 11 interacts with the information processing device 10 through an initial dialogue screen and an expert dialogue screen, thereby causing the information processing device 10 to generate case data and rule trees tailored to the user's individual circumstances.

[0022] The expert terminal 12 can transmit case data of services provided by experts such as medical services, legal services, and accounting / tax services to the information processing device 10. This case data is stored in the case data information 106.

[0023] (System Configuration) Figure 2 is a system configuration diagram of the information processing device 10 according to the present invention. The information processing device 10 comprises a control unit 101, a main memory unit 102, an auxiliary memory unit 103, an IF unit 104, and an output unit 105, which are interconnected by a bus 120 or the like, similar to a general computer. The information processing device 10 contains case data information 106, rule tree information 107, and user-specific persona information 108 in the form of a file / database or the like.

[0024] The control unit 101, also known as the central processing unit (CPU), controls each component of the information processing device 10 and performs data calculations. It also reads various programs stored in the auxiliary storage unit 103 into the main memory unit 102 and executes them. The main memory unit 102, also known as the main memory, stores various received data, computer-executable instructions, and data after calculations performed by those instructions. The auxiliary storage unit 103 is a storage device, such as a hard disk drive (HDD), which stores data and programs for the long term.

[0025] The embodiment shown in Figure 2 describes an embodiment in which the control unit 101, main memory unit 102, and auxiliary storage unit 103 are located inside the same computer. However, in other embodiments, the information processing device 10 can be configured to achieve parallel distributed processing by multiple computers by using multiple control units 101, main memory unit 102, and auxiliary storage unit 103. In another embodiment, it is also possible to set up multiple servers for the information processing device 10, and have multiple servers share a single auxiliary storage unit 103.

[0026] The IF unit 104 acts as an interface (IF) for sending and receiving data with other systems and devices, and provides an interface for receiving various commands and input data (various masters, tables, etc.) from the system operator. The output unit 105 provides a display screen for displaying the processed data and a printing means for printing the data.

[0027] Case data information 106 stores expert elements and case information, which form the basis for generating personas (rule trees). Figure 3 shows an example of the elements and case information of each expert stored in case data information 106, and an example of the rule tree of each expert's persona generated from them. The case data includes the elements of each expert and the case information corresponding to those elements. For example, Figure 3 shows an example of expert A's case data, where the elements and case information are shown. The information processing device 10 can generate case data including elements and case information based on the recognition results of the natural language recognition unit for the input data received from the expert terminal 12 and add it to case data information 106, or it can add case data entered via a case data input screen (not shown) to case data information 106.

[0028] The rule tree information 107 stores information about personas (rule trees) generated based on case data. In the example in Figure 3, the information processing device 10 is shown generating a persona (rule tree) based on each of the expert's case data. The elements exemplified in Figure 3 become elements in the persona, and the case information becomes the element values ​​in the persona.

[0029] Here, the rule tree will be explained with reference to Figure 4. The rule tree has a combination of multiple elements 111 and element values ​​112. Any element (condition attribute) and element value (condition attribute value) in the rule tree may be a target element (conclusion attribute) and a target element value (conclusion attribute value). For example, the element value 112 at the lowest level of the tree may be a target element value (i.e., the conclusion of the persona). The structure of the branch tree and the number of target elements may be determined according to the content of the elements. In this specification, elements 111 and element values ​​112 are sometimes collectively referred to as "elements" (or "nodes"). Element values ​​can be configured to call other elements in the same rule tree or elements in other rule trees, and can also be configured to execute any program (executable file, API).

[0030] When generating (configuring) a rule tree, the information processing device 10 can set the name of the event that generated the rule, the element name, the element value, the information entropy value, and the alias for each element (node). A node may include at least the name of the event that generated the rule, the element type, the element name, the element value, the information entropy value, and the alias. The name of the event that generated the rule indicates the event or personified subject (e.g., expert) that was the source of the rule tree generation. The element type is a flag indicating whether the element type is a condition or a conclusion. The element name is a name that indicates the content of the condition or conclusion. The element value indicates the attribute value of the condition or conclusion. The information entropy value indicates the entropy value of the element calculated by the information processing device 10 when the rule tree was generated. The entropy value can indicate how much information an element contains (also called its "information value"). A high entropy value indicates that it contains a relatively large amount of information, i.e., that its information value is relatively high. Conversely, a low entropy value indicates that it contains a relatively small amount of information, i.e., that its information value is relatively low. Aliases can indicate a value associated with the same meaning when related elements with the same meaning exist in the same or different personas, i.e., a reinterpretation value (synonym). Aliases can have multiple reinterpretation values.

[0031] One embodiment of the present invention can be described in the general context of computer-executable instructions executed by a computer, such as a program module. Generally, a program module includes routines, programs, objects, components, data structures, etc., that perform a specific job or implement a specific abstract data type. Another embodiment of the present invention can also be implemented in a distributed computing environment, where the job may be executed by remote processing units connected through a communication network. In a distributed computing environment, program modules can be located in local computer storage media and remote computer storage media.

[0032] (A process that generates multiple personas tailored to the individual circumstances of the user.) Referring to Figure 5, we will explain the process of generating different multiplexed personas for each user through two types of dialogue. Figure 5 is a flowchart illustrating the process in which a first multiplexed persona is generated through an initial dialogue between the user terminal 11 and the information processing device 10, and then a second multiplexed persona tailored to the user's individual circumstances is generated through expert dialogue.

[0033] In S501, the information processing device 10, in response to access from the user terminal 11, provides the user terminal 11 with the initial dialogue screen (first screen) of the dialogue application. The initial dialogue screen provides an interface in which the user can input or select information about themselves and the content of the services they wish to receive. Note that the content that can be input or selected on this interface differs depending on the field of services provided. The content input or selected on the initial dialogue screen is called initial dialogue information and is used when generating the first multiplexed persona. The user terminal 11 transmits the initial dialogue information to the information processing device 10.

[0034] In S502, when the information processing device 10 receives initial dialogue information from the user terminal 11 via the initial dialogue screen, it analyzes the initial dialogue information. This analysis process may be performed based on whether the initial dialogue information contains predetermined words. If the initial dialogue information is in the form of speech, speech recognition technology may be used to convert the speech data into text, or to grasp meaning, emotion, keywords, etc., from the speech data. Based on the analysis results of the initial dialogue information, the information processing device 10 queries the case data information 106, reads the case data associated with the initial dialogue information, and generates a first expert persona based on each of the case data, as illustrated in Figure 3. The first expert persona is a persona (rule tree) that is generated in common for all users if specific keywords are included in the initial dialogue information.

[0035] The information processing device 10 can merge the case data read from the case data information 106 to generate merged case data as illustrated in Figure 6(a). The merged case data may be a merge of case data from professionals of the same occupation (for example, case data from multiple nurses), or it may be a merge of case data from professionals of different occupations (for example, case data from a doctor, a pharmacist, and a nurse). Based on the merged case data, the information processing device 10 can generate a second expert persona as illustrated in Figure 6(b).

[0036] The first expert persona fulfills the role of an individual expert, while the second expert persona fulfills the role of an expert team. The information processing device 10 can combine the first expert persona and (optionally) the second expert persona to form a first multiplexed persona group. That is, the first multiplexed persona group may consist only of the first expert persona, or it may consist of the first expert persona and the second expert persona. Figure 7(a) shows an example of the first multiplexed persona group, in which three first expert personas and one second expert persona are shown. In another example, the first multiplexed persona group may contain only multiple first expert personas.

[0037] In S503, the information processing device 10, in response to the generation of the first multiplexed persona group, provides the user terminal 11 with an expert dialogue screen (second screen). The expert dialogue screen is an interface for obtaining the final evaluator's (e.g., user's) evaluation for each combination of elements and element values ​​of each persona in the first multiplexed persona group. The information processing device 10 presents the user terminal 11 with a query regarding each combination of elements and element values ​​of each persona in the first multiplexed persona group via the expert dialogue screen, and obtains the user's response from the user terminal 11. The query may be something like, "When XX happens, will you use YY?", where XX corresponds to the content of the element and YY corresponds to the content of the element value. The user's response may be a positive user evaluation (e.g., "Yes", "OK") or a negative user evaluation (e.g., "No", "NG").

[0038] In S504, the information processing device 10, based on the above inquiry and user response, traverses the elements and element values ​​of the rule trees of the first and second expert personas within the first multiplexed persona group to determine which combinations of elements and element values ​​result in a positive or negative user evaluation. This process also allows the information processing device 10 to determine whether identical and similar combinations of elements and element values ​​within each persona in the first multiplexed persona group result in a positive or negative user evaluation. The information processing device 10 can use aliases to determine whether elements and element values ​​are identical or similar, but the determination may also be made based on parameters other than aliases. These combinations of elements and element values ​​may also be target elements and target element values. Based on the combinations of elements and element values ​​that result in a positive user evaluation and the combinations of elements and element values ​​that result in a negative user evaluation, the information processing device 10 updates the case data that formed the basis for generating the first and second expert personas within the first multiplexed persona group. The information processing device 10 generates a first correction expert persona based on the updated case data. The information processing device 10 can merge the updated case data and generate a second correction expert persona based on the merged updated case data.

[0039] The information processing device 10 generates user personas based on combinations of elements and element values ​​that result in positive user evaluations, and combinations of elements and element values ​​that result in negative user evaluations. The information processing device 10 can store updated case data, updated merged case data, user personas, a first corrective expert persona, and a second corrective expert persona in user-specific persona information 108 in association with user information. This allows the information processing device 10 to have case data and expert personas tailored to the individual circumstances of each user.

[0040] In S505, the information processing device 10 can combine the first revision expert persona, the second revision expert persona, and the user persona into a second multiplexed persona group. Figure 7(b) shows an example of the second multiplexed persona group, in which three first revision expert personas, one second revision expert persona, and a user persona are shown. In another example, the second multiplexed persona group may include only multiple first revision expert personas. In this specification, since an example from a medical setting is given, the user persona is shown as a "patient persona."

[0041] As explained below, the second group of multiple personas includes updated personas and user personas, based on the user's responses, resulting in a different persona group for each user.

[0042] (A process that adjusts the element values ​​of the persona based on the individual circumstances of the user.) Figure 8 illustrates the process of adjusting the combination of elements and element values ​​for each persona included in the second persona group. This process can be executed after the user personas are generated in S504.

[0043] In S801, the information processing device 10 determines which combinations of elements and element values ​​result in a positive or negative user evaluation, and adds the value indicating each evaluation to the target element value of the user persona. The user persona includes combinations of elements and element values ​​that result in a positive user evaluation, and combinations of elements and element values ​​that result in a negative user evaluation. The information processing device 10 can add a first value to the target element value of the target element and target element value that can be traced from the combination of elements and element values ​​that result in a positive user evaluation. The information processing device 10 can also add a second value to the target element value of the target element and target element value that can be traced from the combination of elements and element values ​​that result in a negative user evaluation. In Figure 9, when (element, element value) = (c, c1) and (a, a1) are traced by the information processing device 10, it is shown whether each combination of target element and target element value results in a positive or negative evaluation (the tree portion indicated by "*1" in Figure 9). As in this example, the information processing device 10 can determine whether a combination of target element and target element value is positively or negatively evaluated, and can add a positive or negative evaluation to the target element value.

[0044] In S802, the information processing device 10 updates the case data read in S502 based on combinations of elements and element values ​​that result in a positive user evaluation and combinations of elements and element values ​​that result in a negative user evaluation. In this update process, case data associated with combinations of elements and element values ​​that result in a negative user evaluation may be deleted, i.e., updated so that they are not included, or the information entropy value assigned to the element value when the persona is generated may be controlled so that such combinations are less likely to be selected in the dialogue.

[0045] The information processing device 10 generates a first correction expert persona based on the updated case data. The information processing device 10 merges the updated case data and generates a second correction expert persona based on the merged updated case data. Figure 10 shows an example of the first correction expert persona and the second correction expert persona. In Figure 10, combinations where (element, element value) = (b, b2) and (e, e1) are highlighted. The information processing device 10 controls the entropy value of the highlighted element and element value combinations to make them less likely to be selected. In other embodiments of the present invention, a rule tree may be generated so that the highlighted element and element value combinations are not included.

[0046] Here, we will explain the entropy value of the persona. Each node in the persona (rule tree) is associated with an entropy value calculated by the information processing device 10. The entropy value indicates the amount of information (information value) that the node possesses, and a higher entropy value indicates a greater amount of information. For example, a node with a high entropy value, that is, one that contains a lot of information, can be designated as a "high-priority element". Therefore, in the embodiment of the present invention, the information processing device 10 can be configured to control the element selection method by changing the entropy value of each node using parameters and weights based on information on the combination of elements and element values.

[0047] In S803, the information processing device 10 combines the first revision expert persona, the second revision expert persona, and the user persona into a single entity to generate a second group of multiplexed personas, as described in S505.

[0048] According to the method described in Figure 8, the information processing device 10 can identify which combinations of target elements and target element values ​​result in a negative evaluation, thus enabling easy updating of case data. Furthermore, by avoiding the use of combinations that lead to negative evaluations during persona generation, the information processing device 10 can ensure that the ultimately generated first and second revision expert personas are tailored to the individual circumstances of the user.

[0049] Although the principles of the present invention have been described above with reference to exemplary embodiments, various embodiments with modifications in configuration and details can be realized without departing from the spirit of the invention. That is, the present invention can be implemented in the form of, for example, a system, apparatus, method, program, or storage medium. [Explanation of symbols]

[0050] 10 Information Processing Devices 11 User terminals 12 Expert terminals 101 Control Unit 102 Main memory 103 Auxiliary storage 104 IF section 105 Output section 106 Case Data Information 107 Rule Tree Information 108 User-Specific Persona Information 111 elements 112 Element Values

Claims

1. An information processing device comprising a control unit and a storage unit, The memory unit includes a case data DB that stores elements and case information associated with experts, which serve as the basis for generating personas. The control unit, The process involves querying the case data DB based on the initial dialogue information received from the user terminal and reading the case data associated with the initial dialogue information. To generate a first expert persona based on the aforementioned case data read out, To generate a first group of personas, including the first expert persona, Obtain user evaluations from the user terminal for each combination of elements and element values ​​of each persona in the first persona group, Based on the user evaluations, determine whether each combination of elements and element values ​​for each persona in the first persona group is evaluated positively or negatively. Based on the above determination, the retrieved case data is updated, To generate a first revised expert persona based on the updated case data, To generate user personas based on combinations of elements and element values ​​that result in positive user evaluations, and combinations of elements and element values ​​that result in negative user evaluations, To generate a second group of personas, including the first modified expert persona and the user persona, An information processing device configured to perform the following actions.

2. The information processing device according to claim 1, wherein the first expert persona is a user-common persona generated based on specific keywords included in the initial dialogue information.

3. The information processing apparatus according to claim 1, wherein the initial dialogue information includes user information and information about the content of the services the user wishes to receive.

4. A second expert persona is generated based on the merged case data obtained by merging the previously read case data, To generate the first group of personas, including the first expert persona and the second expert persona, A second revised expert persona is generated based on the merged case data generated based on the updated case data, To generate the second group of personas, which includes the first revised expert persona, the second revised expert persona, and the user persona, An information processing apparatus according to claim 1, configured to further perform the following:

5. The information processing apparatus of claim 4, wherein the determination includes determining, based on the user evaluation, which combination of elements and element values ​​results in a positive or negative user evaluation by traversing the same and similar elements and element values ​​in the rule trees of the first expert persona and the second expert persona of the first persona group.

6. The information processing apparatus of claim 5, wherein tracing identical and similar elements and element values ​​is performed based on aliases.

7. Generating user personas based on combinations of elements and element values ​​that result in positive user evaluations, and combinations of elements and element values ​​that result in negative user evaluations, The first value is added to the target element value of the target element and target element value that can be traced from the combination of elements and element values ​​that result in a positive user evaluation, Adding a second value to the target element value of the target element and target element value that can be traced from the combination of elements and element values ​​that result in negative user evaluations. The information processing apparatus according to claim 1, further comprising:

8. The information processing apparatus according to claim 1, wherein updating the read case data based on the determination is performed by deleting the case data associated with combinations of elements and element values ​​that result in a negative user evaluation.

9. An information processing method performed by an information processing device comprising a control unit and a storage unit, The memory unit includes a case data DB that stores elements and case information associated with experts, which serve as the basis for generating personas. The control unit queries the case data DB based on the initial dialogue information received from the user terminal and reads the case data associated with the initial dialogue information. The control unit generates a first expert persona based on the case data read out, The control unit generates a first group of personas, including the first expert persona. The control unit obtains user evaluations from the user terminal for each combination of elements and element values ​​of each persona in the first persona group, The control unit determines, based on the user evaluation, whether each combination of elements and element values ​​of each persona in the first persona group is a positive or negative user evaluation. The control unit updates the read case data based on the determination made, The control unit generates a first revised expert persona based on the updated case data, The control unit generates user personas based on combinations of elements and element values ​​that result in positive user evaluations, and combinations of elements and element values ​​that result in negative user evaluations. The control unit generates a second group of personas, including the first modified expert persona and the user persona. A method for providing this.

10. A program that causes a computer to perform the method described in claim 9.

Images