Information processing device, relevance estimation system, relevance estimation method, and program

Abstract

The aim is to provide technology that enables reliable selection of case studies. [Solution] An information processing device comprising: a request acquisition unit that acquires case request information including search information used for searching for cases; a relevance estimation unit that causes a generating AI to select cases related to the search information, estimates the degree of relevance to the search information for each case, and generates a reason for selection; and a response unit that responds to the source of the case request information with the degree of relevance and the reason for selection for each case selected by the relevance estimation unit.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, a relevance estimation system, a relevance estimation method, and a program.

Background Art

[0002] In fields such as user support, when receiving a user's question, there is a technology for extracting cases related to the question.

[0003] For example, in the related case reference system disclosed in Patent Document 1, a user transmits a question email describing an error message or error log of a trouble related to an information processing system to a responsible terminal. When the responsible terminal transfers the question email to a case accumulation server, the case accumulation server extracts keywords from the question email and searches for cases using the keywords.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Conventionally, as in the technology according to Patent Document 1, in order to search for a case desired by a user, it has been common to perform a process of searching for a character string included in the case using keywords. However, a certain skill is required to represent a desire by a keyword, and in some cases, there is a possibility that the desired case cannot be obtained. Also, when the keyword representing the user's desire is a term with a high degree of abstraction, etc., there may be a mismatch between the obtained case and the user's desire.

[0006] The present invention has been made in view of the above points, and an object thereof is to provide a technology that enables a reliable selection when selecting a case. [Means for solving the problem]

[0007] This application includes several means for solving the above-mentioned problems, some examples of which are as follows.

[0008] To solve the above problems, an information processing device according to one aspect of the present invention is characterized by comprising: a request acquisition unit that acquires case request information including search information used for searching for cases; a relevance estimation unit that causes a generating AI to select cases related to the search information, estimates the degree of relevance with the search information for each case, and generates a reason for selection; and a response unit that responds to the source of the case request information with the degree of relevance and the reason for selection for each case selected by the relevance estimation unit.

[0009] The information processing device may be characterized by comprising a case acquisition unit that acquires past case resolutions, a relevance estimation unit that causes a generating AI to select cases related to the search information from among the case resolutions acquired by the case acquisition unit, and a response unit that includes a predetermined number of the case resolutions with high relevance in the response.

[0010] The aforementioned relevance estimation unit may be characterized by having the generating AI estimate a re-estimated relevance for each case, which is the degree of relevance between the search information and the selection reason and the aforementioned relevance.

[0011] The response unit may be characterized by including the re-estimated relevance in the response as the relevance of the case.

[0012] The information processing device may include a relevance correction unit that corrects the re-estimated relevance using a preset correction rule and obtains the corrected relevance, and the response unit may include the selected examples using the corrected relevance in the response.

[0013] The information processing device may include an example acquisition unit that acquires multiple past examples, and the relevance correction unit may be characterized by correcting the relevance for each of the multiple groups each composed of the multiple examples whose relevance has been estimated.

[0014] The information processing device may be characterized by comprising a case extraction unit that extracts cases for each of the plurality of groups, a relevance estimation unit that causes the generating AI to re-estimate the relevance for each of the cases extracted by the case extraction unit, and a relevance correction unit that performs correction using the relevance re-estimated by the generating AI.

[0015] Furthermore, in order to solve the above problems, another aspect of the present invention relates to a relevance estimation system, which is a computer-based relevance estimation system, characterized in that the computer performs a request acquisition step of acquiring case request information including search information used for searching for cases; a relevance estimation step of causing a generating AI to select cases related to the search information, to estimate the degree of relevance with the search information for each case, and to generate a reason for selection; and a response step of responding to the source of the case request information with the degree of relevance and the reason for selection for each case selected in the relevance estimation step.

[0016] Furthermore, in order to solve the above problems, another aspect of the present invention provides a relevance estimation method that is executed by an information processing device, comprising: a request acquisition procedure for acquiring case request information including search information used for searching for cases; a relevance estimation procedure for causing a generating AI to select cases related to the search information, to estimate the degree of relevance with the search information for each case, and to generate a reason for selection; and a response procedure for responding to the source of the case request information with the degree of relevance and the reason for selection for each case selected in the relevance estimation procedure.

[0017] Also, in order to solve the above problems, a program according to another aspect of the present invention is a program for causing a processing unit of a computer to execute a relevance estimation method, the program including a request acquisition procedure for acquiring case request information including search information used for case search, a relevance estimation procedure for causing a generation AI to select a case related to the search information, estimate a relevance with the search information for each case, and generate a selection reason, and a response procedure for responding with the relevance and the selection reason to the source of the case request information for each case selected in the relevance estimation procedure.

Effect of the Invention

[0018] According to the present invention, it is possible to provide a technique that enables a reliable selection when selecting a case.

[0019] Problems, configurations, and effects other than those described above will be clarified by the description of the following embodiments.

Brief Description of the Drawings

[0020] [Figure 1] It is a diagram showing an example of a functional block of a relevance estimation system in the first embodiment. [Figure 2] It is a diagram showing an example of a data structure of case information. [Figure 3] It is a diagram showing an example of a hardware configuration of a relevance estimation device. [Figure 4] It is a flowchart showing an example of a case selection process in the first embodiment. [Figure 5] It is a diagram showing an example of an overview of a relevance estimation process in a modification of the first embodiment. [Figure 6] It is a flowchart showing an example of a case selection process in a modification of the first embodiment. [Figure 7] It is a diagram showing an example of an overview of a relevance estimation process in the second embodiment. [Figure 8] It is a diagram showing an example of a functional block of a relevance estimation system in the second embodiment. [Figure 9] It is a flowchart showing an example of case selection processing in the second embodiment.

Mode for Carrying Out the Invention

[0021] <First Embodiment>

[0022] Hereinafter, examples of embodiments of the present invention will be described based on the drawings. FIG. 1 is a diagram showing an example of a functional block of a relevance estimation system 1 in the first embodiment. The relevance estimation system 1 in the present embodiment includes a relevance estimation device 10, a case request device 20, and a generation AI system 30. The relevance estimation device 10, the case request device 20, and the generation AI system 30 are communicably connected via a network N.

[0023] As an example of the present embodiment, the case request device 20 is a device operated by a person in charge of providing user support for solving problems that the user has. The relevance estimation device 10 selects cases that have solved similar problems in response to the case request information from the case request device 20 and responds to the case request device 20. The response includes, in addition to the selected cases, a relevance indicating the degree of relevance to the problem, and the reason for selecting the case. The relevance and the reason for selection are output by the case request device 20 together with the selected cases.

[0024] The relevance estimation device 10 is a server computer, PC (Personal Computer), or workstation, etc., and is, for example, a device owned by a business operator providing relevance estimation services. The relevance estimation device 10 comprises a processing unit 110, a storage unit 120, an input unit 130, an output unit 140, and a communication unit 150. The processing unit 110 comprehensively controls the entire relevance estimation device 10. The storage unit 120 stores information necessary for processing by the processing unit 110. The input unit 130 receives information input to the relevance estimation device 10 from input devices connected via the input interface of the relevance estimation device 10. The output unit 140 outputs the information stored in the relevance estimation device 10 to output devices connected via the output interface of the relevance estimation device 10. The communication unit 150 mediates the transmission and reception of information with other information processing devices connected via the network N.

[0025] The processing unit 110 comprises a request acquisition unit 111, a relevance estimation unit 112, a response unit 113, and a case acquisition unit 114. The request acquisition unit 111 acquires case request information, including search information used for searching for cases, from the case request device 20. For example, the search information may be keywords indicating business challenges faced by users handled by the person in charge of the case request device 20, or sentences containing those keywords.

[0026] The relevance estimation unit 112 instructs the generating AI to select cases related to the search information. The relevance estimation unit 112 also instructs the generating AI to estimate the degree of relevance of each case to the search information and to generate the reason for selection. More specifically, the relevance estimation unit 112 generates execution instructions, i.e., prompts, to select cases related to the search information from past cases included in the case information 121 described later. The prompts include execution instructions to estimate the degree of relevance of each selected case to the search information and to generate the reason for selection.

[0027] The relevance estimation unit 112 transmits an execution instruction to the generation AI system 30 and obtains the selected cases, the degree of relevance between those cases and the search information, and the reason for selecting those cases. As an example, the relevance estimation unit 112 causes the generation AI to select a case related to the search information from past resolved cases. The selection of cases includes selecting a predetermined number of cases with high relevance from the cases to be searched. The selection of cases also includes assigning a degree of relevance to all cases to be searched. This is because, when the response unit 113 (described later) outputs the selected cases in order of decreasing relevance, the output of cases with low relevance will have a lower priority, which is effectively equivalent to selecting cases with high priority.

[0028] The response unit 113 responds to the case request device 20, which is the source of the case request information, with the degree of relevance to the search information and the reason for selection for each case selected in response to the case request information. The response unit 113 may include all of the solution cases selected by the relevance estimation unit in its response, or it may include a predetermined number of solution cases with high relevance among the selected cases in its response.

[0029] The case acquisition unit 114 acquires case information about cases that have occurred in the past. For example, the case acquisition unit 114 acquires past resolved cases that are included in the case information 121 and are managed by the provider of the relevance estimation service. The case information acquired by the case acquisition unit 114 is treated as a case to be searched.

[0030] The memory unit 120 stores case information 121. Case information 121 is information about cases acquired by the case acquisition unit 114, but the details will be described later.

[0031] The case request device 20 is an information processing device such as a PC, workstation, smartphone, PDA (Personal Data Assistant), or tablet device. The case request device 20 comprises a processing unit 210, a storage unit 220, an input unit 230, an output unit 240, and a communication unit 250. The processing unit 210 comprehensively controls the entire case request device 20. The storage unit 220 stores information necessary for processing by the processing unit 210. The input unit 230 receives information input to the case request device 20 from input devices connected via the input interface of the case request device 20. The output unit 240 outputs information stored in the case request device 20 to output devices connected via the output interface of the case request device 20. The communication unit 250 mediates the transmission and reception of information with other information processing devices connected via the network N.

[0032] The processing unit 210 includes a case request unit 211. The case request unit 211 transmits case request information, including search information entered by the person in charge, to the relevance estimation device 10. The case request unit 211 outputs the selected case, its relevance, and the reason for selection, which are included in the response transmitted from the relevance estimation device 10, to the output device of the case request device 20.

[0033] The generation AI system 30 utilizes machine learning and natural language processing to generate and output various documents, images, videos, audio, program code, etc., in response to prompts. If an API (Application Programming Interface) is provided, the relevance estimation device 10 can also access the API function to input prompts. The generation AI system 30 can utilize LLMs (Large Language Models) such as OpenAI's CHATGPT® or Google's Gemini. The generation AI system 30 may be built locally, and the relevance estimation device 10 may access it via a LAN (Local Area Network).

[0034] Hereinafter, the generation AI system 30 will be described as a device that generates text data in response to instructions using an LLM. However, the model used by the generation AI system 30 is not limited to an LLM. For example, the generation AI system 30 may select images related to an image input as search information as examples, and generate a document for each example that includes the degree of relevance and the reason for selection.

[0035] Network N is a communication network such as a LAN (Local Area Network), WAN (Wide Area Network), the Internet, or a mobile phone network. Network N may also be a VPN (Virtual Private Network) on a wireless communication network such as a mobile phone network.

[0036] Figure 2 shows an example of the data structure of case information 121. For example, case information 121 includes case ID, case name, case summary, issues, solutions, value provided, customer industry, customer size, service name, characteristics, target business, delivery method, and keywords.

[0037] The Case ID is an identifier that identifies the record of Case Information 121. The Case Name is a string indicating the name of the Case Information (i.e., the record of Case Information 121). The Case Summary is a string indicating a brief description of the case. The Issues are information indicating the specific problems or difficulties faced in the case related to the Case Information. The Resolutions are information indicating the issues that were resolved by the case. The Value Provided is information indicating the value provided to the customer in the case.

[0038] Customer Industry is information indicating the customer's industry. Customer Size is information indicating the customer's size, such as annual sales or number of employees. Service Name is information indicating the name of the customer's service or product. Features are information indicating noteworthy points in the case study. Target Business is information indicating the business content covered by the case study. For example, target business may be "production planning" or "equipment maintenance." Delivery Method is information indicating the specific method used to solve the problem. For example, the delivery method may include the technology implemented or specific procedures. Keywords are information indicating the words used to allow case study searches.

[0039] Note that Case Study Information 121 does not necessarily have to include all of the items shown in Figure 2, and may include other items. For example, Case Study Information 121 may include information such as customer name, project phase indicating the progress of the project, project overview diagram, project scale, information identifying the internal website, introductory materials, interview sheet, collaborating or partner company, registration date for Case Study Information 121, design person, sales person, etc. Also, certain information such as the project overview may be generated by the generation AI when generating selection reasons, etc.

[0040] Figure 3 shows an example of the hardware configuration of the relevance estimation device 10. The relevance estimation device 10 comprises a processor 101, memory 102, storage 103, communication device 104, and a bus 105 connecting the devices. In addition, the relevance estimation device 10 may also include an input device and an output device.

[0041] The processor 101 is a computing device such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit), and it performs processing according to a program recorded in memory 102 or storage 103. In the relevance estimation device 10, processing is performed by the processor 101, which operates according to a program read from memory 102 or storage 103. The processing unit 110, request acquisition unit 111, relevance estimation unit 112, response unit 113, and case acquisition unit 114 each realize their respective functions by having the processor 101 execute a program.

[0042] Memory 102 is a storage device such as RAM (Random Access Memory) or flash memory, and functions as a storage area where programs and data are temporarily read. Storage 103 is a writable and readable storage device. The storage unit 120's functions are realized by either memory 102 or storage 103. Alternatively, the storage unit 120 may be realized by a storage device connected via communication device 104.

[0043] The communication device 104 is an interface for connecting the relevance estimation device 10 to an external device. For example, the communication device 104 uses an antenna that can utilize predetermined radio waves (e.g., 5GHz band, 2.4GHz band, etc.) to establish a connection with other information processing devices using the Wi-Fi standard for wireless communication. The communication unit 150's functions are realized by the communication device 104.

[0044] The processor 101, memory 102, storage 103, and communication device 104 are connected to each other by connecting wires such as a bus 105.

[0045] Each processing unit (request acquisition unit 111, relevance estimation unit 112, response unit 113, and case acquisition unit 114) may be constructed using dedicated hardware (ASIC, GPU, etc.) to implement its respective function. Furthermore, the processing of each processing unit may be executed on a single piece of hardware or on multiple pieces of hardware.

[0046] Since the case request device 20 basically has the same hardware configuration as the relevance estimation device 10, its explanation will be omitted.

[0047] Figure 4 is a flowchart showing an example of the case selection process in the first embodiment. The process in this flowchart begins, for example, when the case request unit 211 of the case request device 20 transmits case request information to the relevance estimation device 10, and the relevance estimation device 10 receives the case request information via the communication unit 150.

[0048] First, the request acquisition unit 111 acquires case request information (step S11). Specifically, the request acquisition unit 111 acquires search information included in the case request information.

[0049] Next, the case acquisition unit 114 acquires case information 121 (step S12). For example, the case acquisition unit 114 acquires past resolved cases from the case information 121. As an example, the case acquisition unit 114 acquires all cases included in the case information 121. As another example, the case acquisition unit 114 may acquire multiple records included in the case information 121 by using the information included in the case request information acquired before the start of this process to extract cases related to the case request information from the case information 121. For example, the case acquisition unit 114 may extract information that identifies a client from the case request information and acquire cases related to that client from the case information 121.

[0050] Next, the relevance estimation unit 112 generates commands to select cases, generate reasons for selection for each case, and estimate the degree of relevance (step S13). Specifically, the relevance estimation unit 112 generates a prompt command to select cases related to the search information from the cases to be searched obtained in step S12. The prompt may include the cases to be searched obtained in step S12, or it may include information specifying the storage location of the cases. The relevance estimation unit 112 includes a command in the prompt to output the degree of relevance between the cases and the search information, and the reasons for selecting the cases. As an example, the relevance estimation unit 112 generates the following prompt.

[0051] (Example of a prompt) Select relevant case studies from the attached case study list based on the entered search information and output them as a list of relevant case studies. The search should include not only term matching but also cases with similar meanings. When outputting, please include the selection reason explaining why each case is relevant to the search information, and a 10-point relevance rating. The output should be presented in descending order of relevance, based on "Case ID," "Case Name," "Selection Reason," and "Relevance Level." "Input Search Information" Please give me some examples related to temperature. "Input Example Information 1" Case ID:SC-1001 Project Name: Remote Maintenance using Temperature and Humidity Sensors and Monitoring Center BPO Project Overview: Combining temperature and humidity sensors with a monitoring center BPO to provide low-cost remote monitoring. Problem: Conventional temperature and humidity sensors had issues such as requiring on-site verification. Solution: By utilizing a monitoring center BPO, real-time remote monitoring and health checks became possible at a low cost. It also became possible to manage multiple systems simultaneously. Keywords: Visualization temperature and humidity sensor "Input Example Information 2" Case ID:SC-1002 Project Name: Weather-Based Traffic Demand Forecasting and Transportation Optimization Project Overview: Predicting traffic volume that changes due to weather conditions and optimizing transportation routes. Problem: Conventional weather-based traffic volume forecasts had large discrepancies, requiring a large margin of error, which led to losses. Solution: Traffic volume was predicted using AI based on weather forecasts, reducing the margin rate to less than 20% of the conventional rate. Furthermore, transportation costs were reduced by optimizing transportation. Keywords: Transportation optimization, Demand forecasting, AI

[0052] In this embodiment, the relevance estimation unit 112 can generate a prompt to select case information related to the search information based on the values ​​included in each item of the case information 121. In this case, it is desirable that the selection reason be created taking into consideration the values ​​included in each item of the case information 121.

[0053] Next, the relevance estimation unit 112 outputs a command and obtains the result (step S14). Specifically, the relevance estimation unit 112 sends the prompt generated in step S13 to the generation AI system 30 and receives the result.

[0054] Next, the response unit 113 generates a response that associates the selection reason with the degree of relevance for a predetermined number of highly relevant cases (step S15). Specifically, the response unit 113 uses the selection results obtained by the degree of relevance estimation unit 112 in step S14 to generate a response that associates the selection reason with the degree of relevance for each predetermined number of highly relevant cases. The response unit may also include the selection results received from the generation AI system 30 directly in the response content. As an example, one case included in the response generated by the response unit 113 is as follows.

[0055] Case ID: AC321J Project Name: Visualization of Line Inspection Results Using Belt Conveyors Project Overview: Real-time graph display of work performance management using cameras. Reason for selection: The objective is to improve operational efficiency in the manufacturing industry, and this may be related to increasing the speed of work on parts that can be transported by conveyor belts. Relevance: 9

[0056] Next, the response unit 113 outputs a response (step S16). For example, the response unit 113 sends the response generated in step S15 to the case request device 20 that sent the case request information. After that, the processing unit 110 terminates the processing of this flowchart.

[0057] In addition, the relevance estimation unit 112 may use the case summary of the case information 121, which has been pre-entered by the person in charge of each case, to have the generating AI generate the selection reasons. The case summary of the case information 121 may also be a summary of the case separately generated by the generating AI, and the relevance estimation unit 112 may use the summary of the case to have the generating AI generate the selection reasons.

[0058] As described above, according to this embodiment, it is possible to select cases that are highly relevant to the user's search criteria, thereby improving convenience. Furthermore, since the degree of relevance of the selected cases to the user's requirements is presented to the user, the user can recognize the depth of the relationship between the requirements and the cases. In addition, since the reason for selecting the cases is displayed, the user can understand why the cases were selected, which not only improves convenience but also increases the reliability of the search results.

[0059] <First Embodiment: Modified Example> Next, the relevance estimation system 1 in a modified version of the first embodiment will be described. The differences from the first embodiment will be described below. For example, due to processing load, there may be an upper limit to the amount of data that the generation AI system 30 can process simultaneously. If the total amount of data for the records of the case information 121 to be searched exceeds the upper limit, the generation AI system 30 divides the case information 121 to be selected into groups and performs distributed processing. In that case, there may be a bias in the degree of relevance for each group, and it may not be possible to obtain appropriate results. The relevance estimation device 10 in this modified version obtains a re-estimated degree of relevance using the degree of relevance estimated for each group.

[0060] Figure 5 shows an example of the overview of the relevance estimation process in a modified version of the first embodiment. In this modified version, the relevance estimation unit 112 of the relevance estimation device 10 accepts the setting of a limit value, which is the upper limit of the amount of data that the generating AI system 30 can process simultaneously. The type of limit value is not limited and may be, for example, an upper limit of the data size, an upper limit of the number of data points, or an upper limit of the number of characters contained in the data.

[0061] As case information 121 to be used for case selection processing, the case acquisition unit 114 acquires 120 "Data A" and 20 "Data B". For example, "Data A" and "Data B" have different data formats because they were generated at different times. In this modified example, the relevance estimation unit 112 can generate different system prompts for "Data A" and "Data B" to compensate for the differences due to the differences in data formats.

[0062] In the example shown in Figure 5, since the data size of "Data A" exceeds the limit, the relevance estimation unit 112 divides the case information 121 of "Data A" into six groups and assigns the case information 121 of "Data B" to one group (step X1). The relevance estimation unit 112 selects cases related to the search information from each of the six groups of "Data A" and the one group of "Data B", estimates the degree of relevance to the search information, and generates a prompt to generate the reason for the selection.

[0063] The relevance estimation unit 112 obtains information about the cases selected for each group as a result of sending each prompt to the generation AI system 30 (step X2). In the example shown in Figure 5, the relevance estimation unit 112 obtains the result that three cases have been selected for each group, but the number of selected cases that the relevance estimation unit 112 obtains from the generation AI system 30 is not limited to this. For example, the relevance estimation unit 112 may obtain all of the cases obtained by the case acquisition unit 114 as selected cases.

[0064] In the example shown in Figure 5, "Data A1," "Data A5," and "Data A9" are selected from the group in the left column. In this figure, "Data A1" is labeled "(Estimated 4 - Expected 6)." This indicates that for "Data A1," the correlation estimated by the generating AI is "4," while the expected correlation is "6." The expected correlation is, for example, the correlation that would be estimated if the generating AI were to estimate correlation for all cases simultaneously. The expected correlation is a theoretical value used for explanation and is not the actual value output by the correlation estimation device 10.

[0065] Furthermore, while Figure 5 shows the selection results for three groups for convenience, the other four groups also have cases selected in the same way as the three groups shown in this figure. In each case shown in the middle section of this figure, the estimated relevance differs from the expected relevance. It is presumed that this is because the generating AI system 30 estimates the relevance for cases belonging to the same group, and therefore the bias in the content of the cases in each group leads to a bias in the estimated relevance. For example, if a certain group has a concentration of cases related to specific search information (e.g., keywords), then when the cases are divided into groups and the relevance of each case is calculated, cases that would have had a high relevance if they had been included in another group may be extracted with a low relevance.

[0066] Therefore, in this modified example, the relevance estimation unit 112 generates a prompt to estimate the re-estimated relevance between the case and the search information using the selection reason and the relevance, and sends it to the generation AI system 30. Since the amount of data for one case is reduced, the number of cases that the generation AI system 30 can process simultaneously increases, and the bias in the re-estimated relevance is suppressed. As a result, the re-estimated relevance for each case is estimated (step X3). Subsequently, the relevance estimation unit 112 treats the re-estimated relevance as the relevance. Figure 5 shows the state in which cases with high re-estimated relevances have been selected. Figure 5 shows the state in which a re-estimated relevance close to the expected relevance has been estimated.

[0067] Figure 6 is a flowchart showing an example of the case selection process in a modified version of the first embodiment. In this modified version, the relevance estimation device 10 has limit values ​​set in advance, for example, before the start of the process shown in this flowchart.

[0068] The processes performed in steps S21 and S22 are the same as those performed in steps S11 and S12 shown in Figure 4, so their explanation is omitted.

[0069] Next, the relevance estimation unit 112 determines whether the acquired case information 121 exceeds a limit value (step S23). Specifically, the relevance estimation unit 112 determines whether the search target case acquired by the case acquisition unit 114 in step S22 exceeds a pre-set limit value.

[0070] If the relevance estimation unit 112 determines that the acquired case information 121 does not exceed the limit value (if the result in step S23 is "NO"), it executes the processes in steps S24 to S27. The processes executed in steps S24 to S27 are the same as those executed in steps S13 to S16 in Figure 4, so their explanation is omitted. After that, the processing unit 110 terminates the process in this flowchart.

[0071] If the relevance estimation unit 112 determines that the acquired case information 121 exceeds a limit value (if the answer is "YES" in step S23), it divides the case information 121 into groups so that it fits within the limit value (step S28).

[0072] Next, the relevance estimation unit 112 generates commands for each group to select cases, generate reasons for selection for each case, and estimate the degree of relevance (step S29). Specifically, the relevance estimation unit 112 generates prompts for each group to select cases related to the search information from among the cases included in that group. The prompts include the records of each case information 121 included in the group as search targets. The relevance estimation unit 112 includes commands in the prompts to output the degree of relevance between the cases and the search information, and the reasons for selecting the cases. As mentioned above, the relevance estimation unit 112 may also output commands to select all the records of each case information 121 included in the group.

[0073] The process performed in step S30 is the same as the process performed in step S25. As a result, the relevance estimation unit 112 obtains information for each group that associates the selected case, the degree of relevance between the case and the search information, and the reason for selection.

[0074] Next, the relevance estimation unit 112 generates a command to re-estimate the relevance using the estimated reason and the relevance (step S31). Specifically, the relevance estimation unit 112 generates a prompt requesting that the re-estimated relevance be estimated for each case using the estimated reason and the relevance obtained in step S29 for all cases selected in that step. Note that the relevance estimation unit 112 does not group the target cases when re-estimating the relevance. That is, all cases extracted in step S17 are subject to re-estimated relevance.

[0075] In addition, the relevance estimation unit 112 may generate a prompt requesting the estimation of re-estimated relevance for a predetermined number of cases with high relevance among the cases selected in step S29. The relevance estimation unit 112 may also generate a prompt to re-select cases using the re-estimated relevance. Furthermore, the relevance estimation unit 112 may generate a prompt requesting the generation of selection reasons again for the re-selected cases.

[0076] The process performed in step S32 is the same as the process performed in step S25, so its explanation is omitted.

[0077] Next, the response unit 113 generates a response that associates the selection reason with the re-estimated relevance for a predetermined number of cases with a large re-estimated relevance (step S33). In other words, the response unit 113 includes the re-estimated relevance as the relevance of the cases in the response. The selection reason included in the response may be a selection reason generated in response to the prompt in step S29, or a selection reason generated in response to the prompt generated in step S31.

[0078] The process performed in step S34 is the same as in step S27, so the explanation is omitted.

[0079] According to the relevance estimation system 1 in this modified example, even when the generating AI divides the cases into groups and performs relevance estimation processing, the influence of bias in the data processed at once on the relevance estimation can be suppressed. In other words, a more accurate relevance can be estimated, and confidence can be gained regarding the selection of cases.

[0080] <Second Embodiment> Next, the relevance estimation system 1 in the second embodiment will be described. The differences from the first embodiment and the modified version of the first embodiment will be described below. In this embodiment, the relevance estimation device 10 obtains the corrected relevance by correcting the re-estimated relevance using a preset correction rule.

[0081] Figure 7 shows an example of the overview of the relevance estimation process in the second embodiment. In this embodiment, the relevance estimation device 10 accepts the setting of a limit value, which is the upper limit of the amount of data that the generation AI system 30 can process simultaneously, similar to the modified example of the first embodiment. If the total number of cases acquired by the case acquisition unit 114 as search targets exceeds the limit value, the relevance estimation device 10 divides the cases into multiple groups so that the total number of cases in each group is less than the limit value. Subsequently, the relevance estimation unit 112 sends prompts to the generation AI system 30 to select cases for each group, estimate the degree of relevance, and generate reasons for selection, and obtains the results.

[0082] Next, the relevance estimation device 10 extracts multiple cases from each of the multiple groups (step Y1). For example, the relevance estimation device 10 determines the number of extractable cases according to a predetermined criterion and extracts the determined number of cases from each of the multiple groups. As an example, if the number of extractable data is "2", the relevance estimation device 10 extracts the case with the highest relevance and the case with the lowest estimate from among the multiple cases included in the group. From the group in the left column of Figure 7, "Data A5" is extracted as the case with the highest relevance, and "Data A1" is extracted as the case with the lowest relevance. From the group in the center column of Figure 7, "Data A100" is extracted as the case with the highest relevance, and "Data A110" is extracted as the case with the lowest relevance. From the group in the right column of Figure 7, "Data B2" is extracted as the case with the highest relevance, and "Data B1" is extracted as the case with the lowest relevance.

[0083] Next, the relevance estimation device 10 uses the selection reasons and relevance of the extracted cases to cause the generating AI system 30 to estimate the re-estimated relevance (step Y2). In the example shown in Figure 7, the relevance estimation device 10 uses the selection reasons and relevance of the extracted cases, "Data A1", "Data A5", "Data A100", "Data A110", "Data B1", and "Data B2", to cause the generating AI system 30 to estimate the re-estimated relevance.

[0084] Next, the relevance estimation device 10 corrects the relevance for each group (step Y3). The relevance estimation device 10 corrects the relevance of each case for each group using a pre-set correction rule and re-estimated relevance. For example, the relevance estimation device 10 corrects the relevance of each case by scaling.

[0085] In the group shown in the center column of Figure 7, the re-estimated relevance of "Data A100" is "10," and the re-estimated relevance of "Data A110" is "8." The relevance estimation device 10 treats the relevance of "Data A100" as being the same as the re-estimated relevance of "10," and the relevance of "Data A110" as being the same as the re-estimated relevance of "8." For "Data A105," which belongs to the group, the relevance estimation device 10 corrects the relevance to "9" by using a correction rule. Note that the description of relevance correction is omitted for the left and right columns of Figure 7.

[0086] Next, the relevance estimation device 10 selects cases using the corrected relevance and outputs the results (step Y4). That is, the relevance estimation device 10 selects cases using the corrected relevance from the cases selected in each group, or from all the searchable cases included in each group, and transmits them to the case request device 20.

[0087] Figure 8 shows an example of the functional blocks of the relevance estimation system 1 in the second embodiment. The processing unit 110 of the relevance estimation device 10 in the second embodiment includes a request acquisition unit 111, a relevance estimation unit 112, a response unit 113, a case acquisition unit 114, as well as a relevance correction unit 115 and a case extraction unit 116.

[0088] The relevance correction unit 115 corrects the re-estimated relevance using a pre-set correction rule to obtain the corrected relevance. The relevance correction unit 115 corrects the relevance for each of the multiple groups, each composed of multiple cases for which relevance has been estimated. The relevance correction unit 115 can correct the relevance using known methods. For example, the relevance correction unit 115 uses a piecewise function to correct the relevance for each case that was not extracted by the case extraction unit 116 described later, based on the re-estimated relevance of the extracted cases.

[0089] The case extraction unit 116 calculates the number of cases that can be extracted from one group using a limit value. The case extraction unit 116 then extracts the calculated number of cases from each group. For example, A limit of "20" data points is set, and if the case acquisition unit 114 acquires "200" searchable cases, then 200 ÷ 20 = "10" groups are generated. The case extraction unit 116 calculates 20 ÷ 10 = "2" as the number of cases that can be extracted from each group. The case extraction unit 116 can use known methods to extract cases from each group. For example, as mentioned above, the case extraction unit 116 extracts the case with the highest relevance and the case with the lowest relevance within each group. Alternatively, the case extraction unit 116 may extract cases so that the relevance of the cases to be extracted is approximately equal. Alternatively, the case extraction unit 116 may extract cases randomly.

[0090] Figure 9 is a flowchart showing an example of the case selection process in the second embodiment. In this embodiment, the relevance estimation device 10 has limit values ​​set in advance, for example, before the start of processing in this flowchart, similar to the modified example in the first embodiment.

[0091] The processes executed in steps S41 to S50 are the same as those executed in steps S21 to S30 in Figure 6, so their explanation is omitted.

[0092] Next, the case extraction unit 116 calculates the number of extractable data points per group using the limit value and the number of groups (step S51). For example, the case extraction unit 116 calculates the number of extractable data points per group by dividing the limit value by the number of groups divided in step S48.

[0093] Next, the case extraction unit 116 extracts a number of cases for each group based on the number of extractable data (step S52). Specifically, for each group obtained in step S48, the case extraction unit 116 extracts the number of cases calculated in step S51 from among the multiple cases included in the group. For example, the case extraction unit 116 extracts the case with the highest degree of relevance and the case with the lowest degree of relevance.

[0094] Next, the relevance estimation unit 112 generates a command to re-estimate the relevance of the extracted cases using the selection reason and the relevance (step S53). Specifically, the relevance estimation unit 112 generates a prompt to cause the generating AI system 30 to re-estimate the relevance using the selection reason and the relevance of the cases extracted in step S52.

[0095] The process performed in step S54 is the same as in step S32 in Figure 6, so the explanation is omitted.

[0096] Next, the relevance estimation unit 112 corrects the relevance of each case for each group using the re-estimated relevance of the extracted cases (step S55). For example, for cases extracted in step S52, the relevance estimation unit 112 determines the corrected relevance to be the re-estimated relevance obtained in step S54. For example, the relevance estimation unit 112 corrects the relevance by scaling the uncorrected relevance of cases not extracted in step S52 based on the relationship between the uncorrected and corrected relevance of the extracted cases. That is, the relevance estimation unit 112 normalizes the relevance of each case based on a predetermined correction rule obtained for each group.

[0097] As a result, the relevance estimation unit 112 obtains the selection reason and the adjusted relevance for each case selected from the cases obtained as search targets in step S42. Alternatively, the relevance estimation unit 112 may treat all cases obtained as search targets as selected and obtain the selection reason and the adjusted relevance for each case. Alternatively, the relevance estimation unit 112 may select cases using the adjusted relevance obtained for all cases obtained as search targets in step S42.

[0098] Next, the response unit 113 generates a response that associates the selection reason with the re-estimated relevance for a predetermined number of cases with a high relevance after correction (step S56). In other words, the response unit 113 includes the cases selected using the corrected relevance in the response.

[0099] The process performed in step S57 is the same as in step S47, so the explanation is omitted.

[0100] In step 52 of this embodiment, if the total number of cases extracted exceeds the limit, the processing unit 110 can perform the processing from grouping in step S48 to relevance correction in step S55 in multiple stages. Specifically, the case extraction unit 116 sets the number of extractable data to 2 in step S51. In step S52, the case extraction unit 116 extracts the number of cases calculated in step S51 from among the multiple cases included in each group obtained in step S48. Next, the case extraction unit 116 determines whether the number of extracted cases exceeds the limit. If the number of extracted cases exceeds the limit, the case extraction unit 116 performs the processing from step S48 to step S52 for the extracted cases.

[0101] For example, if the limit is set to "20" as the maximum number of information per process, and the number of cases to be searched is "360", then the total number of groups is 360 ÷ 20 = "18". The number of extractable data calculated by the case extraction unit 116 is 20 ÷ 18 = 1.11..., which is less than 2. In step S51, the case extraction unit 116 sets the number of extractable data to "2". In step S52, the case extraction unit 116 extracts "2" cases from each of the "18" groups. The case extraction unit 116 determines that the total number of extracted cases, 18 × 2 = "36", exceeds the limit of "20".

[0102] Next, the relevance estimation unit 112 executes step S48, dividing the 36 cases into multiple groups. Since 36 ÷ 20 = 1.8, the relevance estimation unit 112 divides the cases into two groups. After that, the processes from steps S49 to S51 are executed. In step S51, the case extraction unit 116 calculates 20 ÷ 2 = 10 as the number of extractable data. Since the calculated number of extractable data is less than or equal to the limit value of 20, the case extraction unit 116 moves the process to step S53. That is, in step S54, the relevance estimation unit 112 obtains the re-estimated relevance of the extracted cases.

[0103] As described above, if the cases are grouped multiple times, the relevance correction unit 115 performs a multi-stage process in step S55 to correct the relevance using the re-estimated relevance. For example, if the division in step S48 is performed three times, in step S54, the relevance estimation unit 112 obtains the re-estimated relevance of the cases extracted from the group obtained as a result of the third division. The relevance correction unit 115 uses the re-estimated relevance to correct the relevance of the cases included in the group obtained after the third division. This group includes cases extracted from the group obtained as a result of the second division.

[0104] The relevance correction unit 115 corrects the relevance of cases included in a group using the corrected relevance of cases extracted from the group obtained as a result of the second division. By repeating this process, the relevance correction unit 115 obtains normalized relevance for the cases acquired in step S42.

[0105] As described above, the relevance estimation system 1 in this embodiment can obtain highly reliable search results even when there is case information 121 that exceeds the limit value. Furthermore, by selecting cases using a more reliable relevance score, cases that meet the user's requirements can be efficiently selected.

[0106] Although the embodiments of the present invention have been described above, the present invention is not limited to the examples of embodiments described above, and various modifications are included. For example, the examples of embodiments described above are explained in detail to make the present invention easier to understand, and the present invention is not limited to having all the configurations described herein. Furthermore, it is possible to replace a part of the configuration of one example of an embodiment with the configuration of another example. It is also possible to add a configuration of another example to the configuration of one example of an embodiment. Furthermore, it is possible to add, delete, or replace a part of the configuration of one example of each embodiment with a configuration of another example. In addition, some or all of the above configurations, functions, processing units, processing means, etc., may be realized in hardware, for example, by designing them as integrated circuits. Also, the control lines and information lines in the figures are shown only if they are considered necessary for explanation, and do not necessarily show all of them. It can be assumed that almost all of the configurations are interconnected.

[0107] Furthermore, the functional configurations of the relevance estimation device 10, the case request device 20, and the generation AI system 30 described above are classified according to their main processing content for ease of understanding. The present invention is not limited by the way the components are classified or named. As shown above, the configurations of the relevance estimation device 10, the case request device 20, and the generation AI system 30 can be further classified into many more components depending on the processing content. Alternatively, they can be classified so that a single component performs even more processing.

[0108] For example, the relevance estimation device 10 may function as a case request device 20. In that case, the relevance estimation device 10 can cause the generating AI system 30 to select cases in response to search information transmitted from the user's terminal device. [Explanation of Symbols]

[0109] 1: Relevance estimation system, 10: Relevance estimation device, 20: Case request device, 30: Generation AI system, 110-210: Processing unit, 111: Request acquisition unit, 112: Relevance estimation unit, 113: Response unit, 114: Case acquisition unit, 115: Relevance correction unit, 116: Case extraction unit, 120-220: Memory unit, 121: Case information, 130-230: Input unit, 140-240: Output unit, 150-250: Communication unit, 211: Case request unit, 101: Processor, 102: Memory, 103: Storage, 104: Communication device, 105: Bus

Claims

1. A request acquisition unit that acquires case request information including search information used for searching for cases, A relevance estimation unit that causes the generating AI to select examples related to the search information, estimates the degree of relevance of each example to the search information, and generates the reason for selection, An information processing device comprising: a response unit that responds to the source of the case request information with the degree of relevance and the reason for selection for each case selected by the degree of relevance estimation unit.

2. An information processing apparatus according to claim 1, It is equipped with a case acquisition unit that acquires past case examples. The relevance estimation unit causes the generation AI to select a case related to the search information from among the solution cases acquired by the case acquisition unit. The response unit is characterized in that it includes a predetermined number of highly relevant solution cases in the response.

3. An information processing apparatus according to claim 1, The information processing device is characterized in that the relevance estimation unit causes the generating AI to estimate a re-estimated relevance for each case, which is the degree of relevance between the search information and the selection reason and the relevance.

4. An information processing apparatus according to claim 3, The information processing device is characterized in that the response unit includes the re-estimated relevance in the response as the relevance of the case.

5. An information processing apparatus according to claim 3, The system includes a correlation correction unit that corrects the re-estimated correlation using a pre-set correction rule and obtains the corrected correlation, The response unit is characterized in that it includes the example selected using the corrected relevance in the response.

6. An information processing device according to claim 5, It is equipped with a case acquisition unit that acquires multiple past cases, The information processing device is characterized in that the correlation correction unit corrects the correlation for each of the multiple groups, each composed of the multiple examples for which the correlation has been estimated.

7. An information processing apparatus according to claim 6, The system includes a case extraction unit that extracts the aforementioned cases for each of the aforementioned multiple groups, The correlation estimation unit causes the generating AI to re-estimate the correlation for each of the cases extracted by the case extraction unit. The information processing device is characterized in that the correlation correction unit performs a correction using the correlation re-estimated in the generated AI.

8. A computer-based relevance estimation system, The aforementioned computer, A request acquisition step to acquire case request information that includes search information used to search for cases, The generation AI is instructed to select examples related to the search information, to estimate the degree of relevance of each example to the search information, and to generate the reason for selection. A relevance estimation system characterized by performing a response step in which, for each of the cases selected in the relevance estimation step, the system responds to the source of the case request information with the degree of relevance and the reason for selection.

9. A method for estimating relevance performed by an information processing device, A request retrieval procedure for obtaining case request information, which includes search information used to search for cases, A correlation estimation procedure in which the generating AI is instructed to select examples related to the search information, to estimate the degree of relevance of each example to the search information, and to generate the reason for selection, A method for estimating relevance, characterized by comprising a response procedure for responding to the source of the case request information with the degree of relevance and the reason for selection for each case selected in the relevance estimation procedure.

10. A program that causes a computer's processing unit to execute a correlation estimation method, A request retrieval procedure for obtaining case request information, which includes search information used to search for cases, A correlation estimation procedure in which the generating AI is instructed to select examples related to the search information, to estimate the degree of relevance of each example to the search information, and to generate the reason for selection, A program characterized by causing the program to execute a response procedure that, for each of the cases selected in the relevance estimation procedure, responds to the source of the case request information with the degree of relevance and the reason for selection.

Images