Information processing device, information output method, and information output program

Abstract

This information processing device (10) comprises: a reception unit (15A) that receives text describing a question from a user terminal; an extraction unit (15C) that extracts an event corresponding to the question on the basis of schedule-related information (13A) in which events for which an action is scheduled by a user belonging to a specific group are collected, and the text describing the question; and an output unit (15E) that outputs information on the event obtained as the extraction result to the user terminal.

Description

Information processing device, information output method, and information output program

[0001] The present invention relates to an information processing device, an information output method, and an information output program.

[0002] In order to promote knowledge sharing and problem solving in groups such as organizations and communities, there are support systems that support the formation of connections with people who have knowledge by using the group's knowledge information and questions from users as input information.

[0003] Japanese Patent Application Laid-Open No. 2017-219899

[0004] Yudai Nagata, Takayuki Ito, Toramatsu Shintani, "Proposal of a Knowledge Market Based on Reputation in Knowledge Sharing Systems," Proceedings of the 66th National Conference, Vol. 2004, No. 1, pp. 293-294, March 9, 2004

[0005] However, in the above-mentioned support system, if knowledge corresponding to the user's question is not accumulated, it is not possible to suggest connections with people who have the knowledge, which means that it is not possible to provide information that will help the user solve their problem or share information.

[0006] Therefore, an object of the present invention is to provide an information processing device, an information output method, and an information output program that can realize information sharing among users or providing information that contributes to problem solving.

[0007] In order to solve the above-mentioned problems and achieve the object, the information processing device of the present invention has a reception unit that receives text in which a question is written from a user terminal, an extraction unit that extracts an event corresponding to the question based on schedule information that collects events in which users belonging to a specific group plan to act and the text in which the question is written, and an output unit that outputs information related to the event obtained as a result of the extraction to the user terminal.

[0008] According to the present invention, it is possible to provide information that contributes to information sharing among users or problem solving.

[0009] FIG. 1 is a block diagram showing an example of the functional configuration of an information processing device. FIG. 2 is a schematic diagram illustrating a support system according to a conventional technology. FIG. 3 is a schematic diagram illustrating the problems of a support system according to a conventional technology. FIG. 4 is a schematic diagram illustrating one aspect of a problem-solving approach. FIG. 5 is a schematic diagram (1) illustrating an information output function. FIG. 6 is a schematic diagram (2) illustrating the information output function. FIG. 7 is a schematic diagram (3) illustrating the information output function. FIG. 8 is a diagram illustrating an example of information collected by a collection unit. FIG. 9 is a schematic diagram illustrating the relationship between input and output. FIG. 10 is a diagram (1) illustrating an example of input and output of an extraction unit and a calculation unit. FIG. 11 is a diagram (2) illustrating an example of input and output of an extraction unit and a calculation unit. FIG. 12 is a schematic diagram illustrating an example in which the information output function is applied. FIG. 13 is a diagram (1) illustrating an example of a prompt. FIG. 14 is a diagram (1) illustrating an example of an LLM output. FIG. 15 is a diagram (2) illustrating an example of a prompt. FIG. 16 is a diagram (2) illustrating an example of an LLM output. Fig. 17 is a diagram (3) showing an example of a prompt. Fig. 18 is a diagram (3) showing an example of an LLM output. Fig. 19 is a flowchart showing the procedure for information output processing. Fig. 20 is a diagram showing an example of a hardware configuration.

[0010] Hereinafter, an information processing device, an information output method, and an information output program according to the present application (hereinafter referred to as "embodiments") will be described with reference to the accompanying drawings. Note that the embodiments merely illustrate one example or one aspect, and the following description does not limit the structure, action, function, properties, characteristics, methods, uses, etc. according to the present disclosure.

[0011] <Overall Configuration> Fig. 1 is a block diagram showing an example of the functional configuration of an information processing device 10. Fig. 1 shows the information processing device 10 that provides an information output function for extracting an event corresponding to a user's question from schedule-related information that collects events for which users belonging to a specific group are planning to act, and outputting the extraction result.

[0012] The following is merely an example of a use case in which the group of users to whom the information output function is provided is an organization such as a company, but this does not preclude application to other groups, for example, communities such as a region.

[0013] In one aspect, the information processing device 10 may be realized by a server device. For example, the information processing device 10 can provide the information output function as a cloud service by executing platform as a service (PaaS) type middleware or software as a service (SaaS) type application.

[0014] 1, the information processing device 10 can be communicatively connected to a user terminal 30 and a platform 50 via a network NW. For example, the network NW may be any type of communication network, such as the Internet or a local area network (LAN), whether wired or wireless.

[0015] The user terminal 30 is a terminal device used by a user who receives the information output function. For example, the user terminal 30 may be realized by any computer, such as a personal computer, a smartphone, a tablet terminal, or a wearable terminal.

[0016] The platform 50 is a platform on which various tools used by users belonging to the above group run, including, for example, business communication tools such as schedulers, as well as social networking services (SNSs), blogs, chat services, email services, groupware, and business software.

[0017] Although the above example illustrates the information output function being provided as a cloud service, the present invention is not limited to this. For example, the information output function may be provided on-premise. Furthermore, the information output function may be packaged as one function of a tool used by users belonging to the group.

[0018] Although the above-described information output function is implemented as a client-server system, the present invention is not limited to this. For example, the information output function may be provided as a standalone function by causing an application running on the user terminal 30 to execute processing corresponding to the information output function on the user terminal 30.

[0019] <Example of Prior Art> For example, as explained in the example of the background art above, a support system has been proposed that connects users with each other with the aim of sharing knowledge within a group (company, region).

[0020] Such support systems use AI and other tools to search for staff who can answer users' questions based on knowledge accumulated in systems such as user profiles, Q&A systems, and chat posts. Q&A stands for "Question and Answer," and AI stands for "Artificial Intelligence."

[0021] FIG. 2 is a schematic diagram illustrating a support system according to a conventional technique. For example, FIG. 2 shows an example in which the support system receives a question such as "I want to know about the latest version of System A" from user terminal 30B used by user B. Upon receiving such a question, the support system searches a knowledge database (DB) in which knowledge is accumulated. In this case, knowledge that user A has knowledge about updates to System A is registered in the knowledge DB via user terminal 30A used by user A. Therefore, the support system suggests a connection with user A in response to the question from user terminal 30B.

[0022] <One aspect of the problem> However, in the above support system, if knowledge corresponding to the user's question is not accumulated, it is not possible to suggest connections with people who have the knowledge, which means that it is not possible to provide information that will help the user solve their problem.

[0023] FIG. 3 is a schematic diagram illustrating a problem with a conventional assistance system. For example, FIG. 3 shows an example in which the assistance system receives a question such as "I want to know about the latest version of System B" from user terminal 30B used by user B. In this case, the knowledge DB does not register knowledge about System B from user terminals used by other users. For example, the knowledge DB does not store knowledge such as profiles of users with the ability to answer questions from user B or posts related to questions from user B. This makes it difficult for the assistance system to suggest connections with other users in response to questions from user terminal 30B.

[0024] In addition to the above support systems, other technologies have been proposed, such as adding meta-information to knowledge information to increase the relevance between questions and knowledge, and providing incentives for registering knowledge to promote the accumulation of knowledge, but these technologies also have similar issues.

[0025] <One aspect of the problem-solving approach> Therefore, in this embodiment, a problem-solving approach is adopted that provides an information output function that connects users who face a common problem, from the aspect of supporting effective connections between users even when there are no users with knowledge.

[0026] Figure 4 is a schematic diagram illustrating one aspect of the problem-solving approach. As shown in Figure 4, by proposing a match between user B and other users who are interested in the problem they have or who may have the same problem in the future, a win-win connection is realized. The users targeted for matching here are not limited to individuals, but can also refer to groups containing multiple users formed within the above-mentioned group, such as members or parts of a team, department, office, or section. By including proposals for already formed groups of multiple people as targets for matching, the psychological hurdle of group formation can be lowered.

[0027] In this way, the information output function according to this embodiment estimates people or groups who will need knowledge in the future, even if knowledge has not been accumulated, and promotes problem solving and know-how sharing through connections.

[0028] FIG. 5 is a schematic diagram (1) illustrating the information output function. For example, FIG. 5 shows an example in which the information output function according to this embodiment receives an inquiry from user terminal 30B used by user B, including the question, "The printer won't connect." Upon receiving such a question, the information output function according to this embodiment extracts future events in which the user will face the same challenges as the user from schedule information that collects events planned by users belonging to the group, such as the team's shared calendar and chat / conference information. Then, in response to the inquiry from user terminal 30B, the information output function outputs the extracted events themselves and a proposal for matching with users (including, of course, groups) participating in the events.

[0029] An example of such an event extraction operation is shown in Fig. 6. Fig. 6 is a schematic diagram (2) for explaining the information output function. Fig. 6 also shows an example in which the information output function according to this embodiment receives an inquiry including the question "The printer won't connect" from user terminal 30B used by user B.

[0030] As shown in FIG. 6, the event extraction in the information output function described above may be realized by a trained or fine-tuned large-scale language model, so-called LLM (Large Language Model) 3, by way of example only.

[0031] The LLM3 receives the question "The printer won't connect" from user B and the schedule information described above. With this question and schedule information input, the LLM3 extracts the event "Planned Material Distribution" based on inferences such as "Since a material distribution is scheduled, the user will likely encounter problems using the printer." The information output function described above can then output the event "Planned Material Distribution" in response to an inquiry from user terminal 30B, or output a proposal for matching with team T, which will be participating in the event "Planned Material Distribution." While an example of proposing a match with team T to user terminal 30B used by user B has been given here, it goes without saying that a match with user B can also be proposed to the user terminal 30B used by team T.

[0032] In this way, even if knowledge does not exist (even if the supply and demand situation does not hold), it is possible to provide candidates for connecting with individuals or groups related to that knowledge, taking into account future actions and situations.

[0033] Therefore, the information output function according to this embodiment can provide information that contributes to information sharing among users or problem solving.

[0034] Furthermore, the information output function of this embodiment does not rely solely on the group to carry out future activities; it allows other teams to connect through common interests, creating synergy. For example, in the example of Figure 6, this not only benefits user B, who makes the inquiry, in information sharing and problem solving, but also team T. In this way, rather than relying on one side helping the other, each side can propose valuable collaborative actions, building a mutually beneficial relationship and creating synergy. As described above, even without accumulated knowledge, individuals and groups can experience effective interpersonal connections, achieving a balanced collaborative work environment between individuals and groups.

[0035] Furthermore, while Figure 6 shows an example in which the information input to LLM3 is plan-related information, it is also possible to combine information on behavioral regulations that regulate the user's behavior and information on the surrounding environment that facilitates the extraction of events and input it to LLM3.

[0036] 7 is a schematic diagram (3) illustrating the information output function. Fig. 7 also shows an example in which the information output function according to this embodiment receives an inquiry from user terminal 30B used by user B, including the question "The printer won't connect."

[0037] As shown in FIG. 7 , LLM3 receives a question from user B, "The printer won't connect," as well as behavioral rule information and surrounding environment information. With this question, schedule information, behavioral rule information, and surrounding environment information input, LLM3 extracts Team T's event "Scheduled Meeting" based on inferences such as, "Because they've been requested to print materials for the meeting, they'll likely encounter problems using the printer." Furthermore, LLM3 extracts Person A's event "Delivery Processing" based on inferences such as, "Because they'll likely take action to print a delivery note during delivery processing, they'll likely use the printer." Based on this, the information output function can output the events "Scheduled Meeting" and "Delivery Processing" in response to an inquiry from user terminal 30B. Additionally, the information output function can output a matching proposal with Team T, who will be participating in the "Scheduled Meeting" event, and a matching proposal with Person A, who will be participating in the "Delivery Processing" event. Here, we have given an example of proposing a match with Team T or Mr. A to the user terminal 30B used by User B, but it goes without saying that a match with User B can also be proposed to the user terminal 30 used by Team T or Mr. A.

[0038] In this way, by combining schedule-related information with behavioral regulation-related information and surrounding environment information and inputting the combined information into the LLM 3, it is possible to improve the accuracy of event extraction.

[0039] <Configuration of Information Processing Device 10> Next, the functional configuration of the information processing device 10 that provides the above-described information output function will be described. Fig. 1 schematically illustrates blocks related to the information output function of the information processing device 10. As shown in Fig. 1, the information processing device 10 has a communication control unit 11, a storage unit 13, and a control unit 15. Note that Fig. 1 merely illustrates a selection of functional units related to the above-described information output function, and the information processing device 10 may also be provided with functional units other than those illustrated.

[0040] The communication control unit 11 is a functional unit that controls communication with other devices such as the user terminal 30 and the platform 50. In one embodiment, the communication control unit 11 can be realized by a network interface card such as a LAN card. In one aspect, the communication control unit 11 receives an inquiry including text describing a question from the user terminal 30, or outputs a response to the inquiry, such as a proposal for matching with an event or a user participating in the event, to the user terminal 30.

[0041] The storage unit 13 is a functional unit that stores various types of data. In one embodiment, the storage unit 13 may be realized by internal, external, or auxiliary storage of the information processing device 10. For example, the storage unit 13 stores schedule information 13A, behavioral rule information 13B, and surrounding environment information 13C. Note that the schedule information 13A, behavioral rule information 13B, and surrounding environment information 13C will be described later together with the scenes in which each data is referenced or registered.

[0042] The control unit 15 is a functional unit that performs overall control of the information processing device 10. For example, the control unit 15 may be realized by a hardware processor. As shown in FIG. 1 , the control unit 15 includes a reception unit 15A, a collection unit 15B, an extraction unit 15C, a calculation unit 15D, and an output unit 15E. Note that the control unit 15 may also be realized by hardwired logic or the like.

[0043] The reception unit 15A is a processing unit that receives various information from the user terminal 30. In one aspect, the reception unit 15A can receive an inquiry from the user terminal 30, including text in which a question is written in a natural language. Hereinafter, the text in which a question is written in a natural language may be referred to as "question text." At the stage of receiving the inquiry in this manner, it is assumed that the identification information and account of the user who receives the information output function described above have been identified by user authentication, such as login authentication.

[0044] The collection unit 15B is a processing unit that collects information used in event extraction executed by the extraction unit 15C described below. In one aspect, the collection unit 15B collects the latest schedule information, behavioral rule information, and surrounding environment information when receiving an inquiry from the user terminal 30. Note that, here, an example is given in which the latest schedule information, behavioral rule information, and surrounding environment information are collected every time an inquiry is received from the user terminal 30, but the schedule information, behavioral rule information, and surrounding environment information may also be collected at a fixed cycle or at a regular time.

[0045] Fig. 8 is a diagram showing an example of information collected by the collection unit 15B. As shown in Fig. 8, examples of information collected by the collection unit 15B include schedule information, behavioral rule information, and surrounding environment information.

[0046] Among these, the plan-related information refers to information such as plans, event announcements, tasks, schedules, hopes, and expectations of users belonging to the group. In one embodiment, the collection unit 15B collects, as plan-related information, events that users belonging to the group plan to participate in from the platform 50, which stores primary sources such as scheduler information, diaries, and posts on social media, or forecasts. Such specifications regarding upcoming activities planned by users are referred to as "plan-related information." For example, as shown in FIG. 8 , plan-related information can be described in the form of a shared team calendar (including information such as event details, date, time, location, and participants) or social media posts (including free-form information about events, schedules, dates, and geographical information). These expression formats are natural languages, and are expressed in structured formats such as Markdown and programming language formats, or unstructured formats such as sentences and conversations, and are expressed in computer-interpretable digital formats.

[0047] Behavioral prescriptive information refers to information such as behavioral patterns and behavioral predictions derived from past records of users belonging to the group. In one embodiment, the collection unit 15B collects information prescribing the behavior of users belonging to the group from a platform 50 that stores primary sources such as manuals, procedures, rules, work histories, past user behavior records on social media, and behavioral information including selection results for multiple behavioral options. Such behavioral specifications, such as behavior following a set procedure or predictions of an individual selecting similar behavior from multiple behavioral options based on past experience and information, are referred to as "behavioral prescriptive information." For example, behavioral prescriptive information can be described by a meeting moderator's procedure manual (shown in FIG. 8 ), a device setup manual (including operating procedures, operating methods, operating conditions, etc.), or past user behavior records on social media (including comments on overseas business trips, scenic photos, and the possibility of predicted behaviors other than the purpose of the business trip). These expressions are expressed in computer-interpretable digital formats, including natural language and images.

[0048] Surrounding environment information refers to information representing the attributes, identifications, and characteristics of named entities (named entities) that appear in information entered by users belonging to the above group, such as proper nouns, such as individuals, groups, and the devices they use. In one embodiment, the collection unit 15B collects information such as attributes, identifications, and characteristics of the named entities from a platform 50 that stores primary sources of attributes, identifications, and characteristics of named entities such as individuals, groups, and the devices they use. Such information, which expresses the differentiation and characteristics of individuals, groups, and devices from other devices of the same type, is called "surrounding environment information." For example, in the example shown in FIG. 8 , for example, if a user refers to a "printer" when querying the system, the information describes the location of the printer used by the individual, the printer's update history, and the purpose for which the printer was used by multiple users (e.g., to print meeting materials, to print invoices, etc.). These expressions are expressed in natural language and in a digital format that can be interpreted by a computer.

[0049] When collecting these three pieces of information from the platform 50, the database can be searched by keyword search, vector search, or by RAG (Retrieval-Augmented Generation), which combines these searches with a large-scale language model. In this case, the collection unit 15B can index the database based on the large-scale language model as knowledge data and generate answers in the form of sentences.

[0050] The schedule information, behavioral regulation information, and surrounding environment information collected by the collection unit 15B in this manner are stored in the storage unit 13 as schedule information 13A, behavioral regulation information 13B, and surrounding environment information 13C.

[0051] The extraction unit 15C is a processing unit that extracts events corresponding to the question based on the schedule information 13A, which collects events in which users belonging to the above group are planning to act, and the question text received by the reception unit 15A.

[0052] In one embodiment, the extraction unit 15C generates a prompt in which the question text and the schedule information 13A are embedded. For example, the prompt may include elements of input data such as the question text and the schedule information 13A. The prompt may also include an instruction element for extracting, from the schedule information 13A, events of a user who faces a problem that is the same as or similar to the problem corresponding to the question text. The prompt may also include an output format element that specifies the type and format of the output of the LLM 3. By inputting such a prompt to the LLM 3, the LLM 3 can be caused to output the event extraction results.

[0053] The calculation unit 15D is a processing unit that calculates the degree of association between an event and a question for each event obtained as the extraction result by the extraction unit 15C.

[0054] In one embodiment, the calculation unit 15D generates a prompt in which the event extraction results by the extraction unit 15C and the behavioral rule information 13B are embedded. For example, the prompt may include elements of input data such as the question text, the event extraction results, and the behavioral rule information 13B. The prompt may also include an instruction element that, by referencing the behavioral rule information 13B, determines a higher relevance level for each event obtained as the extraction result, as the relevance between the event and the problem corresponding to the question text increases. The prompt may also include an output format element that specifies the type and format of output from the LLM 3. By inputting such a prompt to the LLM 3, the LLM 3 can be caused to output the relevance level for each event.

[0055] Here, the input / output relationships of the collection unit 15B, extraction unit 15C, and calculation unit 15D will be described. FIG. 9 is a schematic diagram illustrating the input / output relationships. As shown in FIG. 9, the collection unit 15B inputs information stored in the database of the platform 50 to the LLM3. For example, in the case of plan-related information 13A, information such as emails, SNS posts, or schedules stored in the database of a server device that provides services such as email, SNS, and business communication tools is input to the LLM3. As a result, the collection unit 15B collects the plan-related information 13A, behavioral regulation information 13B, and surrounding environment information 13C output by the LLM3.

[0056] The schedule information 13A, behavioral rule information 13B, and surrounding environment information 13C collected in this manner are embedded in a prompt for event extraction by the extraction unit 15C and a prompt for relevance calculation by the calculation unit 15D.

[0057] FIG. 10 is a diagram (1) showing an example of input / output of the extraction unit 15C and the calculation unit 15D. For example, FIG. 10 shows an example of input / output in simple mode. In the example shown in FIG. 10, the extraction unit 15C inputs a prompt in which question text and schedule-related information 13A are embedded to the LLM 3, thereby obtaining an event extraction result output by the LLM 3. The calculation unit 15D then inputs the event extraction result by the extraction unit 15C and a prompt in which behavioral rule-related information 13B is embedded to the LLM 3, thereby obtaining a relevance calculation result output by the LLM 3.

[0058] 11 is a diagram (2) showing an example of inputs and outputs of the extraction unit 15C and the calculation unit 15D. For example, FIG. 11 shows an example of inputs and outputs of an improved mode that is an improvement of the simple mode shown in FIG.

[0059] In the example shown in FIG. 11 , the extraction unit 15C inputs a prompt to the LLM 3, in which attribute information, behavioral rule information 13B, and surrounding environment information 13C are embedded in addition to the question text and schedule information 13A. For example, the attribute information may be the affiliation, job title, or rank of the user belonging to the group. Embedding such attribute information in the prompt makes it easier to extract events of users with attributes similar to those of the user asking the question. Furthermore, embedding the behavioral rule information 13B and surrounding environment information 13C in the prompt makes it possible to evaluate in more detail the similarity between the problem faced by the user asking the question and the problem faced by users participating in the event to be extracted. As a result, it is possible to extract events of users facing problems more similar to those faced by the user asking the question.

[0060] The calculation unit 15D then inputs to the LLM 3 a prompt in which the schedule information 13A and the surrounding environment information 13C are embedded, in addition to the event extraction result by the extraction unit 15C and the behavioral guideline information 13B, to obtain the calculation result of the relevance output by the LLM 3. By embedding the schedule information 13A and the surrounding environment information 13C in the prompt, it is possible to evaluate in more detail the similarity between the problem faced by the user asking the question and the events obtained as the extraction result. As a result, it is possible to assign a high relevance to events that are more similar to the problem faced by the user asking the question.

[0061] 1 , the output unit 15E is a processing unit that executes output control for the user terminal 30. In one aspect, the output unit 15E outputs, to the user terminal 30, information about the event obtained as a result of extraction by the extraction unit 15C.

[0062] In one aspect, the output unit 15E can display a list of events obtained as a result of extraction by the extraction unit 15C in any format, for example, in a table format, on the user terminal 30. In this case, the output unit 15E can also display the relevance calculated by the calculation unit 15D in association with each event obtained as a result of extraction by the extraction unit 15C.

[0063] In another aspect, the output unit 15E can display on the user terminal 30 a list of users participating in the event obtained as a result of extraction by the extraction unit 15C in any format, for example, in a table format. In this case, the output unit 15E can also display the relevance calculated by the calculation unit 15D for each user in association with the list. Furthermore, the output unit 15E can also display recommendations suggesting matching with users participating in the event obtained as a result of extraction by the extraction unit 15C. In this case, the output unit 15E can not only display potential matched users on the user terminal used by the inquiring user, but also suggest matching with the inquiring user on the user terminal 30 of the potential matched users. In this case, if the users participating in the event obtained as a result of extraction are a group, a function such as sending to a common platform to which the group belongs or a notification function for common calendar information may be used.

[0064] <Specific Example> Next, the operation of the information output function according to this embodiment will be described using a specific example. FIG. 12 is a schematic diagram illustrating an example in which the information output function is applied. FIG. 12 illustrates an example in which User B is unable to use the company's printing machine and seeks assistance from the information processing device 10, which provides the status output function described above. In this case, as shown in FIG. 12, the user terminal 30B inputs an inquiry to the information processing device 10, including the text "The printer won't connect," in which the question is written in natural language. Upon receiving this inquiry, the information processing device 10 extracts from the schedule information 13A events in which the user will face the same problem in the future and outputs the extraction results. As a result, the information processing device 10 outputs events that are likely to require the use of the printing machine in the near future, such as weekly meetings, delivery processing, and progress reports, as well as individuals and groups who will participate in those events.

[0065] More specifically, when the query shown in FIG. 12 is input to the information processing device 10, the extraction unit 15C generates the prompt shown in FIG. 13 as an example of a prompt used for event extraction. FIG. 13 is a diagram (1) illustrating an example of a prompt. As shown in FIG. 13, the prompt includes input data elements such as the question text "The printer won't connect," the surrounding environment information 13C described in the heading ##, and the planned information 13A. Furthermore, the prompt includes an instruction element for extracting events of users who are experiencing the same or similar problem as the problem corresponding to the question text "The printer won't connect" from the planned information 13A. Furthermore, the prompt includes an instruction element for outputting the basis for the event extraction. Furthermore, the prompt includes an output format element for causing the LLM3 to output in JSON (Java Script Object Notation) format.

[0066] By inputting the prompt illustrated in FIG. 13 into the LLM 3 in this manner, the LLM 3 can output the event extraction results shown in FIG. 14. FIG. 14 is a diagram (1) showing an example output from the LLM 3. As shown in FIG. 14, three events, "Weekly Project Meeting," "Delivery Processing," and "Weekly Progress Report," are extracted from the events included in the schedule information 13A provided by the prompt illustrated in FIG. 13. These three events may be extracted along with the time and location of each event. Furthermore, the basis for extracting the three events is that scheduled events related to the printer on the 8th floor of the main building were extracted based on the expectation that User B plans to use the printer on the 8th floor of the main building.

[0067] The calculation unit 15D then generates the prompt shown in FIG. 15 as an example of a prompt used to calculate the relevance. FIG. 15 is a diagram (2) showing an example of a prompt. As shown in FIG. 15, the prompt includes input data elements such as the question text "The printer won't connect" and the behavioral rule information 13B, which includes the event extraction results by the extraction unit 15C and internal chat information described in headings ##. Furthermore, the prompt includes an instruction element for determining a higher relevance level for each event obtained as the extraction result by the extraction unit 15C, with reference to the behavioral rule information 13B, as the relevance between the event and the problem corresponding to the question text "The printer won't connect" increases. Furthermore, the prompt includes an instruction element for outputting the basis for calculating the relevance level. Furthermore, the prompt includes an output format element for causing the LLM3 to output the results in a table format.

[0068] By inputting the prompt shown in FIG. 15 into the LLM 3, the LLM 3 can output the relevance calculation results shown in FIG. 16 . FIG. 16 is a diagram (2) showing an example output of the LLM 3. As shown in FIG. 16, an example output of the LLM 3 is shown, which includes relevance evaluations on a three-level scale (high, medium, and low) for the three events given in the prompt shown in FIG. 15 : "Weekly Project Meeting," "Delivery Processing," and "Weekly Progress Report," as well as the basis for calculating the relevance. For example, it can be seen that the events with the highest relevance to the problem faced by user B are "Weekly Progress Report," "Delivery Processing," and "Weekly Project Meeting." Thus, the output of the LLM 3 shown in FIG. 16 clearly shows that the relevance of events with a high probability of using a printer is highly evaluated, while the relevance of events with a low probability of using a printer is low.

[0069] 15 illustrates the in-house chat information as an example of the behavioral rule information 13B, but other information may also be used. For example, the calculation unit 15D may generate a prompt that uses an in-house manual in addition to the in-house chat information as another example of the behavioral rule information 13B.

[0070] Figure 17 is a diagram (3) showing an example of a prompt. In Figure 17, parts that differ from the prompt shown in Figure 15 are highlighted in bold. The prompt shown in Figure 17 differs from the prompt shown in Figure 15 in that an in-house manual, "Manual for Delivery Processing," has been added as input data.

[0071] By inputting the prompt illustrated in FIG. 17 into the LLM3, the LLM3 can output the relevance calculation results shown in FIG. 18 . FIG. 18 is a diagram (3) showing an example output from the LLM3. Like the output from the LLM3 shown in FIG. 16 , the output from the LLM3 shown in FIG. 18 includes relevance ratings on a three-level scale—“high,” “medium,” and “low”—for each of the three events given in the prompts shown in FIG. 17 : “Weekly Project Meeting,” “Delivery Processing,” and “Weekly Progress Report,” along with the basis for calculating the relevance ratings. In this case, as with the output from the LLM3 shown in FIG. 16 , it can be seen that the events with the highest relevance to the problem faced by user B are “Delivery Processing,” “Weekly Progress Report,” and “Weekly Project Meeting.” Thus, like the output from the LLM3 shown in FIG. 16 , the output from the LLM3 shown in FIG. 18 clearly evaluates the relevance of events with a high probability of using a printer, while evaluating the relevance of events with a low probability of using a printer low.

[0072] <Processing Flow> Next, the processing flow of the information processing device 10 according to this embodiment will be described. Fig. 19 is a flowchart showing the procedure of information output processing. As shown in Fig. 19, the reception unit 15A receives an inquiry including a question text in which the question is written in natural language from the user terminal 30 (step S101). Next, the collection unit 15B collects plan-related information, behavioral regulation-related information, and the like from the platform 50 (step S102).

[0073] Then, the extraction unit 15C extracts an event corresponding to the question based on the schedule-related information 13A collected in step S102 and the question text received in step S101 (step S103).

[0074] Thereafter, the calculation unit 15D calculates the degree of association between each event and each question for each event obtained as a result of extraction in step S103 (step S104).Then, the output unit 15E outputs the events obtained as a result of extraction in step S103 and the degree of association calculated for each event in step S104 to the user terminal 30 (step S105), and the process ends.

[0075] <Summary> As described above, the information processing device 10 according to this embodiment extracts events corresponding to the user's question from the schedule information 13A, which collects events in which users belonging to a specific group are planning to act, and outputs the extraction results.

[0076] Therefore, the information processing device 10 according to this embodiment can provide information that contributes to information sharing or problem solving for users. In other words, the information processing device 10 according to this embodiment can achieve synergy by connecting with other teams through common interests, rather than relying on the group itself to carry out future activities. In this way, rather than relying on one party helping the other or being helped, each party can propose valuable joint actions, thereby building a mutually beneficial relationship and achieving synergy. As described above, even without accumulated knowledge, individuals and groups can experience effective interpersonal connections, achieving a balanced collaborative state between individuals and groups.

[0077] <Exercising Creativity> The matters described in this embodiment, such as specific examples of question texts and prompts, and the types of LLMs, are merely examples and can be changed. Furthermore, the order of the processes in the flowchart described in this embodiment can be changed or some processes can be skipped within a consistent range.

[0078] <System> The information including the processing procedures, control procedures, specific names, various data, and parameters shown in the above documents and drawings can be changed as desired unless otherwise specified. For example, any one or more of the functional units of the reception unit 15A, collection unit 15B, extraction unit 15C, calculation unit 15D, and output unit 15E of the information processing device 10 may be configured as separate devices.

[0079] Furthermore, the components of each device shown in the figure are functional concepts and do not necessarily have to be physically configured as shown. In other words, the specific form of distribution and integration of each device is not limited to that shown. In other words, all or part of them can be functionally or physically distributed and integrated in any unit depending on various loads, usage conditions, etc. Note that each configuration may also be a physical configuration.

[0080] Furthermore, all or any part of the processing performed by the illustrated device may be realized by a program executed by a hardware processor such as an MPU (Micro-Processing Unit) or a CPU (Central Processing Unit), or may be realized as hardware using wired logic.

[0081] <Hardware> Next, an example of the hardware configuration of the information processing device 10 described in this embodiment will be described. For example, the information processing device 10 can be implemented by installing a program that realizes the functions of the information processing device 10 on a computer. For example, the information processing device 10 can be implemented by having the computer execute the program, which is provided as package software or online software. The computer referred to here includes desktop or notebook personal computers, rack-mounted server computers, and the like. Furthermore, the computer also includes smartphones, mobile phones, PHS (Personal Handyphone System) and other mobile communication terminals, as well as PDAs (Personal Digital Assistants). The functions of the information processing device 10 may also be implemented on a cloud server.

[0082] An example of a computer that executes the above program (information output program) will be described using Fig. 20. As shown in Fig. 20, the computer 1000 includes, for example, a memory 1010, a CPU 1020, a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.

[0083] The memory 1010 includes a read-only memory (ROM) 1011 and a random access memory (RAM) 1012. The ROM 1011 stores a boot program such as a basic input / output system (BIOS). The hard disk drive interface 1030 is connected to a hard disk drive 1090. The disk drive interface 1040 is connected to a disk drive 1100. A removable storage medium such as a magnetic disk or optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to a mouse 1110 and a keyboard 1120, for example. The video adapter 1060 is connected to a display 1130, for example.

[0084] 20, the hard disk drive 1090 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. The storage unit 13 described in the above embodiment is provided in, for example, the hard disk drive 1090 or the memory 1010.

[0085] Then, the CPU 1020 reads out the program module 1093 and the program data 1094 stored in the hard disk drive 1090 into the RAM 1012 as necessary, and executes the above-mentioned procedures.

[0086] The program module 1093 and program data 1094 relating to the information output program are not limited to being stored in the hard disk drive 1090, but may be stored in a removable storage medium, for example, and read by the CPU 1020 via the disk drive 1100. Alternatively, the program module 1093 and program data 1094 relating to the program may be stored in another computer connected via a network such as a LAN or a WAN (Wide Area Network), and read by the CPU 1020 via the network interface 1070.

[0087] REFERENCE SIGNS LIST 10 Information processing device 11 Communication control unit 13 Storage unit 13A Plan-related information 13B Behavioral regulation-related information 13C Surrounding environment information 15 Control unit 15A Reception unit 15B Collection unit 15C Extraction unit 15D Calculation unit 15E Output unit 30 User terminal 50 Platform

Claims

1. An information processing device comprising: a reception unit that receives text in which a question is written from a user terminal; an extraction unit that extracts events corresponding to the question based on schedule information that collects events in which users belonging to a specific group are planning to act and the text in which the question is written; and an output unit that outputs information related to the events obtained as a result of the extraction to the user terminal.

2. The information processing device described in claim 1, further comprising a calculation unit that calculates the relevance between the event and the question for each event obtained as the extraction result, and the output unit further outputs the relevance in association with the event.

3. An information output method executed by an information processing device, comprising: a receiving step of receiving text in which a question is written from a user terminal; an extraction step of extracting an event corresponding to the question based on schedule information in which events in which users belonging to a specific group are planning to act are collected and the text in which the question is written; and an output step of outputting information related to the event obtained as a result of the extraction to the user terminal.

4. An information output program for causing a computer to execute the following steps: a receiving step of receiving text in which a question is written from a user terminal; an extraction step of extracting an event corresponding to the question based on schedule information in which events in which users belonging to a specific group plan to act are collected and the text in which the question is written; and an output step of outputting information related to the event obtained as a result of the extraction to the user terminal.

Images