Information providing device, information providing method, and information providing program

Abstract

An information providing device (10) includes a scope generation unit (121), an information collection unit (123), and an answer determination unit (124). The scope generation unit (121) generates a plurality of scopes relating to a problem on the basis of an inquiry message related to the problem. The information collection unit (123): generates, for at least one of the plurality of scopes, a question sentence in which the situation of the problem is questioned, and a target user of the question sentence; transmits the question sentence to the user; and acquires an answer to the question sentence. The answer determination unit (124) determines an operation for solving the problem on the basis of the answer.

Description

Information providing device, information providing method, and information providing program

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

[0002] There is a support task that assists users in using IT equipment (e.g., PCs). The support staff (operators) interact with users and may check information outside of the system log, such as the details of the operations performed by the users.

[0003] Tatsuaki Kimura, Megumi Takeshita, Tsuyoshi Toyono, Masahiro Yokota, Ken Nishimatsu, and Tatsuya Mori, "Network Fault Detection and Cause Analysis Technology Using Syslog and SNS Analysis," NTT Technical Journal, July 2013, https: / / journal.ntt.co.jp / backnumber2 / 1307 / files / jn201307020.pdf

[0004] However, with conventional techniques, there are cases where support work cannot be carried out efficiently.

[0005] For example, in support work, there are cases where multiple users make inquiries about different matters. Also, there are cases where different users are responsible for checking information and issuing instructions depending on the situation. Generally, an operator performs support work by himself while managing all information at once. This places a heavy burden on the operator in support work.

[0006] In response to this, it is conceivable to have a system automatically perform tasks such as checking information, collecting information, and issuing instructions for response. However, this would require the creation of detailed scripts, which would increase the costs of maintaining the system and creating the scripts.

[0007] In order to solve the above-mentioned problems and achieve the objectives, the information providing device is characterized by having a scope generation unit that generates multiple scopes related to a problem based on an inquiry message related to the problem, an information collection unit that generates a question inquiring about the situation of the problem and a user targeted by the question for at least one of the multiple scopes, sends the question to the user, and obtains an answer to the question, and an answer determination unit that determines an action to take to solve the problem based on the answer.

[0008] According to the present invention, support work can be carried out efficiently.

[0009] FIG. 1 is a diagram illustrating an example of the configuration of an information providing device according to a first embodiment. FIG. 2 is a flow diagram illustrating a processing flow of the information providing device. FIG. 3 is a diagram illustrating problem isolation in an example. FIG. 4 is a diagram illustrating an overview of processing of the information providing device in an example. FIG. 5 is a diagram illustrating an example of user information. FIG. 6 is a diagram illustrating an example of a prompt for generating a scope. FIG. 7 is a diagram illustrating an example of a generated scope. FIG. 8 is a diagram illustrating an example of a prompt. FIG. 9 is a diagram illustrating an example of generated information collection destinations and hearing statements. FIG. 10 is a diagram illustrating an example of a prompt for answer determination. FIG. 11 is a diagram illustrating an example of generated answer determination results. FIG. 12 is a diagram illustrating an example of generated answer determination results. FIG. 13 is a diagram illustrating an example of a summary. FIG. 14 is a diagram illustrating a conventional support operation. FIG. 15 is a diagram illustrating a support operation according to an embodiment. FIG. 16 is a diagram illustrating an example of a computer that executes an information providing program.

[0010] Hereinafter, embodiments of an information providing device, an information providing method, and an information providing program according to the present application will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described below.

[0011] [First Embodiment] The first embodiment aims to improve efficiency by automating support operations, such as customer support and technical support, that deal with user troubles. In this embodiment, the support operations are assumed to be troubleshooting operations that occur in a certain group or team. The information providing device of this embodiment acts as a facilitator in the support operations, reducing the burden on operators.

[0012] 1 is a diagram showing an example of the configuration of an information providing apparatus according to the first embodiment. The information providing apparatus 10 accepts an inquiry from a user. The inquiry includes text (sentences), images, and the like.

[0013] The information providing device 10 collects information from multiple users (including users other than the user who made the inquiry) based on inquiries and information stored in a database. The information providing device 10 can use large language models (LLMs) when collecting information. The information providing device 10 collects information until a trigger occurs, and then outputs the results of problem isolation based on the collected information.

[0014] Before describing the details of the information providing device 10, the overall business flow including the processing of the information providing device 10 will be described with reference to Fig. 2. Fig. 2 is a flow diagram illustrating the flow of processing of the information providing device.

[0015] The user in FIG. 2 is a user who makes an inquiry. For example, the user is an employee of a company. The related party is a target of information collection. For example, the related party is an employee of the same company as the user. The operator is a person in charge of support work.

[0016] 2, the user sends a message such as "My PC cannot connect to the Internet" to the information providing device 10 (step S100). The information providing device 10 receives the message (step S101).

[0017] The information providing device 10 refers to the manual DB 131 and generates a plurality of questions from the message (step S102).The information providing device 10 sets the generated scope (step S103).

[0018] The information providing device 10 acquires user information from the user DB 132 (step S104). The information providing device 10 generates a question or instruction message with the information collection destination (step S105). The information providing device 10 transmits the question or instruction message to the information collection destination (user, related person, or operator) (step S106).

[0019] The information collection destination receives the question or instruction (step S107) and transmits a response (step S108). The information providing device 10 receives the response (step S109) and determines whether to reset the scope, continue information collection, or terminate it.

[0020] If the information providing device 10 determines to end the process, it completes the determination (step S111) and transmits the determination result to the user (step S112).The user receives the determination result (step S113).

[0021] Returning to Figure 1, the information providing device 10 has an input unit 111, an input unit 112, an output unit 113, an output unit 114, a scope generating unit 121, a scope setting unit 122, an information collecting unit 123, an answer determining unit 124, a determination result providing unit 125, a manual DB 131, and a user DB 132.

[0022] The input unit 111 accepts information input from the terminal 20 held by the user via an interface such as a dedicated form or a chat system. The terminal 20 is an input interface such as a PC, smartphone, or tablet terminal. The input unit 111 allows input of text, images, etc. In addition to information (inquiry) input by the user, the input unit 111 also accepts input of information automatically assigned by the input interface (identifier information indicating which terminal the information was sent from, and flag information indicating whether it is a new inquiry or an ongoing inquiry).

[0023] The input unit 111 passes the input information to the scope generation unit 121. Furthermore, the input unit 112 accepts input of the same information as that input to the input unit 111 from the terminal 30 held by the relevant person. The input unit 112 passes the input information to the answer determination unit 124.

[0024] The scope generation unit 121 processes the inquiry received from the input unit 111. The scope generation unit 121 uses the LLM to isolate the problem (scope) based on the input information and the manual to be referenced. The scope generation unit 121 acquires the manual.

[0025] The scope generation unit 121 combines the input information (inquiry by text or image, etc.), a manual (or operating procedure manual) registered in the manual DB 131, and a prompt that gives instructions to the LLM, and inputs the combined information to the LLM.

[0026] The scope generation unit 121 outputs to the LLM the problem that has been isolated to resolve the cause of the inquiry, checkpoints for the problem, and the user group that can investigate the problem. The scope generation unit 121 passes the output result of the LLM to the scope setting unit 122.

[0027] The scope setting unit 122 sets which scope to focus on when collecting information, based on the information received from the scope generation unit 121 or the answer determination unit 124. For example, the scope setting unit 122 sets one scope corresponding to the problem to be solved from among the separated problems.

[0028] The scope setting unit 122 may set the scope according to pre-registered conditions (e.g., a small unit of unsolved problems among the divided problems) or may set the scope by estimation using a machine learning model including LLM. The scope setting unit 122 passes the setting result to the information collecting unit 123.

[0029] The information collection unit 123 determines from whom information collection will be performed for the scope received from the scope setting unit 122, and also generates questions or instructions (hereinafter referred to as hearing statements) for eliciting information.

[0030] The information collection unit 123 combines the scope information, user information, and prompt received from the scope setting unit 122 and inputs the combined information to the LLM. As a result, the information collection unit 123 generates information on the users (stakeholders) from whom information is to be collected and a hearing statement.

[0031] The user information is a group of information including the ID of the person in charge and the attribute information of the user (department name, floor, working status, skills), etc. The information collection unit 123 transmits the generated question or instruction via the output unit 113 to the terminal 30 of the person concerned from whom information is to be collected.

[0032] The answer determination unit 124 determines the next process based on the information input to the input unit 112. The input unit 112 accepts input of answers to questions or instructions and status (such as an inquiry being handled) from the parties concerned from whom information is collected. The answers are input in the form of text, images, etc.

[0033] The answer determination unit 124 combines the answer (text, image, etc.) received from the input unit 112 with the prompt and inputs it to the LLM, which then causes the LLM to select the next action.

[0034] For example, the options for the next operation are, as described in FIG. 2, scope resetting by the scope setting unit 122, continuation of information collection by the information collection unit 123, or termination processing by the determination result providing unit 125.

[0035] For example, the scope setting unit 122 may select to reset the scope when the problem inquired about is likely to be a different problem from the currently set scope, or when additional investigation of the problem is required.

[0036] For example, continuation of information collection by the information collection unit 123 is selected when collection of more detailed information about the problem is to be continued.

[0037] The termination process by the determination result providing unit 125 is selected when the information collection is sufficiently satisfied (for example, when it is determined that the problem in the current scope is high as a result of problem isolation).

[0038] The answer determination unit 124 outputs the selected option and the reason why the option was selected in a format that can be interpreted by a human.

[0039] The determination result providing unit 125 summarizes the information collection history and generates a sentence to provide the determination result to the user. When the answer determining unit 124 selects termination processing by the determination result providing unit 125, the determination result providing unit 125 combines the information collected from the relevant parties with the prompt and inputs it to the LLM. As a result, the determination result providing unit 125 causes the LLM to generate a summarized sentence and the determination result.

[0040] The output unit 113 and the output unit 114 transmit the information output by the information providing device 10 to the user through an external interface. The output unit 113 and the output unit 114 may transmit data through a SaaS service or the like using an API or the like, or may transmit data using a dedicated form or email or the like.

[0041] [About LLM] The LLM in this embodiment will be described. LLM is a large-scale deep learning language model pre-trained on a huge amount of data. Recently, various LLMs have been proposed, including GPT (registered trademark) (OpenAI), Bard (registered trademark) (Google), and LLaMa (Meta). These models are based on a mechanism called a transformer (specifically, using only the decoder). A transformer is a technology for understanding context by grasping how words and phrases in a sentence are related. In basic operation, an LLM receives text (a prompt) and outputs appropriate text through a transformer.

[0042] After the initial question is input, LLM's internal operation predicts the first character (strictly speaking, a token) that will appear next from the input sentence, adds one character to the end of the prompt, and predicts the next character, repeating this process until a termination condition is met, and then outputs the results up to that point.

[0043] For this reason, the way questions and instructions are posed in the input text (prompt) is extremely important, and by making adjustments, more appropriate answers can be obtained from the LLM. These LLM capabilities are used for predicting the continuation of a sentence, as well as for summarizing sentences, translating, question answering, and text search.

[0044] The information providing device 10 may be a cloud-based or in-house server. The large-scale language model server and ticket server may also be cloud-based or in-house servers. A known technology such as GPT-API can be used for the cloud-based LLM.

[0045] The LLM of this embodiment satisfies the following conditions, for example: Input and output are performed at least in text format. Learning and parameters are sufficiently large-scale. Reference values: data volume is approximately 570 GB (gigabytes), number of parameters is approximately 175 billion (ChatGPT) Learning data: information on basic security concepts, best practices, common security threats, encryption, network security, computer security, data security, etc. (e.g., Wikipedia material on basic concepts of security cyber-attacks (https: / / en.wikipedia.org / wiki / Cyberattack)) A certain degree of real-time output is possible (for example, a minimum response time of one minute per process is the target (ChatGPT)). Able to accept a number of input tokens (as a guideline, a maximum of 2048 input tokens and output tokens can be limited.)

[0046] [Example] An example based on the first embodiment will be described. In this example, a use case is troubleshooting of in-house IT equipment, and it is assumed that a message (inquiry) stating "my PC cannot connect to the in-house network" is sent to the information providing device 10 from a user who has made an inquiry.

[0047] The user making the inquiry is a company employee. The user comes to work and attempts to connect to the company network (Wi-Fi (registered trademark)) using his or her personal company PC, when a network connection failure occurs on the PC. The user, who has a certain level of IT skills, attempts to resolve the issue himself or herself (reviewing network settings, rebooting the PC), but is unable to resolve the issue, so he or she makes an inquiry to the information providing device 10.

[0048] Fig. 3 is a diagram for explaining how to isolate a problem in the embodiment, and Fig. 4 is a diagram for explaining an outline of the processing of the information providing device in the embodiment.

[0049] As shown in Figure 4, the information providing device 10 (1) responds to an inquiry from a user, (2) isolates the problem, (3) sends a hearing to the relevant parties, and (4-1, 4-2) obtains their responses, and (5) contacts an operator as necessary.

[0050] The user information of the user who made the inquiry is as shown in Fig. 5. Fig. 5 is a diagram showing an example of user information. User information 501 is the user information of user "Yamada Taro".

[0051] The user information may include data entered by the user into the inquiry input form when sending an inquiry message, or data automatically provided by the system (such as a work clock-in / work clock-out log).

[0052] For example, when you enter information into a new inquiry form, the information "inquirer" will be added to the inquiry message, and information such as "relationship to self: user, contact: xxx.mail.com" will also be added to the input form.

[0053] The scope generation unit 121 divides the problem and generates a scope. The scope generation unit 121 causes the LLM to generate scopes in order of increasing problem impact range. The scope generation unit 121 also causes the LLM to read manuals, etc., and output the division of the problem. The scope generation unit 121 also causes the LLM to infer interview items and survey subjects.

[0054] In the scope generation, multiple scopes are generated at different granularities depending on the scale of the users. For example, as shown in Figure 3, the problems that form the scopes include a "device problem" that is limited to one user, and a "network problem" that is limited to multiple users.

[0055] Fig. 6 is a diagram showing an example of a prompt for generating a scope. Fig. 7 is a diagram showing an example of a generated scope. The scope generation unit 121 inputs the prompt 502 in Fig. 6 to the LLM and obtains the output 503 in Fig. 7.

[0056] The first line in Figure 6, "I can't connect to the company network," is the content of an inquiry from a user. "Scope" in Figures 6 and 7 is text that represents a scope. "Investigator" is the person involved in the scope. In Figure 7, the scopes are listed in ascending order of scale, such as "individual PC," "department / floor range," and "company-wide."

[0057] The scope setting unit 122 sets the scopes in ascending order of scale. First, the scope setting unit 122 sets "individual PC" as the scope.

[0058] The information collection unit 123 inputs the prompt 504 in FIG. 8 into the LLM to set the information collection destination and generate a hearing sentence. FIG. 8 is a diagram showing an example of the prompt. As shown in FIG. 8, the prompt uses the set scope ("individual PC") information to allow the LLM to select multiple appropriate information collection destination users from the user DB 132. The user DB 132 contains user information for the users included in "enem" in the prompt 504. The information collection destination may be the user who made the inquiry, or a related party other than the user who made the inquiry.

[0059] Furthermore, the information collection unit 123 may select an information collection destination using a general user search function without using the LLM.

[0060] Here, it is assumed that the output 505 shown in Fig. 9 is obtained from the LLM. Fig. 9 is a diagram showing an example of the generated information gathering destination and hearing sentence. The output 505 shows that the information gathering destination ("who") is "Yamada Taro." Furthermore, the "hearing" in the output 505 shows the hearing sentence.

[0061] The answer determination unit 124 sends a message to "Yamada Taro" to collect information based on the generated hearing sentence.

[0062] For example, the answer determination unit 124 sends a question message saying, "Please check the status of the device's Wi-Fi or wired LAN connection," and receives a response from the user "Yamada Taro" such as, "Wi-Fi is running normally."

[0063] Based on the response, the answer determination unit 124 inputs a prompt 506 shown in Fig. 10 to the LLM, causing it to output the answer determination result. Fig. 10 is a diagram showing an example of a prompt for answer determination.

[0064] 11 is a diagram showing an example of generated answer determination results. Output 507 is an answer determination result for prompt 506. Output 508 is an example of an answer determination result for another prompt.

[0065] The answer determination unit 124 determines the next action based on the answer determination result. "Result" in outputs 507 and 508 corresponds to the next action. As described above, the next action is either a scope reset by the scope setting unit 122 ("Result" is "1"), continuation of information collection by the information collection unit 123 ("Result" is "2"), or termination processing by the determination result providing unit 125 ("Result" is "3"). Note that "Reason" is auxiliary information that administrators and the like refer to in order to make LLM processing transparent.

[0066] 12 shows another example of the answer determination result obtained by the answer determination unit 124. FIG. 12 is a diagram showing an example of the generated answer determination result. The "result" in output 509 is "3", indicating that the next operation is termination processing by the determination result providing unit 125. The "result" in output 510 is "1", indicating that the next operation is scope resetting by the scope setting unit 122.

[0067] Furthermore, if the next operation is termination processing by the determination result providing unit 125, the determination result providing unit 125 finally generates a summary to explain the process up to that point to a user (e.g., an operator) who can solve the problem, sends the generated summary to the terminal of the person involved, and ends the process. The summary is generated by inputting all input and output sentences up to that point into the LLM, giving instructions to summarize, generating an answer from the LLM, and sending it to the person involved, and then ending the process. Figure 13 is a diagram showing an example of a summary. The "solver" in output 511 of Figure 13 is a user who can solve the problem.

[0068] Other Examples The first embodiment can be applied to various use cases other than troubleshooting of in-house IT devices.

[0069] For example, the first embodiment can be applied to troubleshooting IT equipment in schools, where even teachers who are not familiar with IT will naturally be able to troubleshoot IT equipment problems.

[0070] For example, if the screen of an IT device used by a student goes blank during class, the teacher inquires of the information providing device 10. The information providing device 10 collects information by asking the teacher who made the inquiry which device (shared PC or individual tablet) the problem occurred on, and if it was a shared PC, by asking the teacher who taught the previous lesson. Furthermore, if the IT device is a student's individual tablet, the information providing device 10 collects information by asking whether the problem also occurred on other students in the same grade at the same time.

[0071] For example, the first embodiment can be applied to a task of resolving troubles caused by spam emails, thereby reducing the burden of interviewing multiple people to identify the extent of the impact of spam emails.

[0072] Consider a case where a user opens a received email and discovers that the email is a phishing email. In this case, in response to an inquiry from the recipient, the information providing device 10 automatically conducts interviews with users around the recipient to gather information.

[0073] For example, the information providing device 10 collects information by asking the recipient about the actions they took in response to the phishing email (whether they clicked on the URL or not), or by asking users around the recipient or users who have a relationship with the recipient (for example, users who have a history of exchanging emails with the recipient) whether they have received similar phishing emails.

[0074] Based on the results of information collection, the information providing device 10 automatically determines whether the damage caused by the phishing email can be resolved on-site or whether a specialized maintenance person should be contacted, and provides a summary of the results to the recipient and related parties.

[0075] Effect of the First Embodiment The scope generation unit 121 generates multiple scopes related to a problem based on an inquiry message related to the problem. The information collection unit 123 generates a question asking about the situation of the problem and a user targeted by the question for at least one of the multiple scopes, sends the question to the user, and acquires an answer to the question. The answer determination unit 124 determines an action to take to solve the problem based on the answer. For example, generating a user means generating information that identifies the user.

[0076] This automates tasks such as problem isolation and information gathering, which were a burden on operators. As a result, this embodiment allows support tasks to be carried out efficiently.

[0077] The scope generation unit 121 generates multiple scopes related to a problem by inputting a prompt to the LLM instructing it to generate multiple scopes with different impact ranges. For example, the "individual PC," "department / floor range," "company-wide scale," and "remote access" shown in Figure 7 are multiple scopes with different impact ranges. This helps solve problems at lower cost by, for example, solving them in ascending order of impact range.

[0078] The answer determination unit 124 determines which of the following actions should be performed: to reset the scope using the scope generation unit 121, to acquire more answers (information collection) using the information collection unit 123, or to perform termination processing. This makes it possible to resolve information shortages and set a more effective scope.

[0079] The determination result providing unit 125 creates a summary of the answers acquired by the information collecting unit 123 and outputs the created summary, thereby enabling necessary information to be quickly conveyed to an operator or the like.

[0080] Fig. 14 is a diagram illustrating a conventional support operation. Fig. 15 is a diagram illustrating a support operation according to an embodiment. As shown in Fig. 14, conventionally, when an operator receives an inquiry from a user, the operator personally interviews relevant parties and gathers information from those around them.

[0081] In contrast, according to this embodiment, as shown in Fig. 15, the system (information providing device 10) collects information on behalf of the operator, thereby reducing the burden on the operator. Furthermore, the information providing device 10 notifies the operator when advanced decision-making or judgment is required.

[0082] [Program] In one embodiment, the information providing device 10 can be implemented by installing an information providing program that executes the above-described processes as package software or online software on a desired computer. For example, by executing the above-described information providing program on an information processing device, the information processing device can function as the information providing device 10. The information processing device referred to here includes desktop and notebook personal computers. Other information processing devices also include smartphones, tablet terminals, and the like.

[0083] The information providing device 10 may also be implemented as a server device that provides services related to the above-described processing to a client terminal device used by a user. For example, the server device may be implemented as a server device that receives an inquiry about a problem as input and outputs a summary of solutions to the problem. In this case, the server device may be implemented as a web server or as a cloud that provides services related to the above-described processing through outsourcing.

[0084] 16 is a diagram showing an example of a computer that executes an information provision program. The computer 1000 includes, for example, a memory 1010 and a CPU 1020. The computer 1000 also includes 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.

[0085] The memory 1010 includes a ROM (Read Only Memory) 1011 and a RAM (Random Access Memory) 1012. The ROM 1011 stores, for example, a boot program such as a BIOS (Basic Input Output System). 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, for example, a mouse 1110 and a keyboard 1120. The video adapter 1060 is connected to, for example, a display 1130.

[0086] The hard disk drive 1090 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. That is, the program that defines each process of the information providing device 10 is implemented as a program module 1093 in which computer-executable code is written. The program module 1093 is stored, for example, in the hard disk drive 1090. For example, the program module 1093 for executing processes similar to those of the functional configuration of the information providing device 10 is stored in the hard disk drive 1090. Note that the hard disk drive 1090 may be replaced with an SSD (Solid State Drive).

[0087] Furthermore, setting data used in the processing of the above-described embodiment is stored as program data 1094, for example, in the memory 1010 or the hard disk drive 1090. The CPU 1020 then reads the program module 1093 or the program data 1094 stored in the memory 1010 or the hard disk drive 1090 into the RAM 1012 as necessary, and executes the processing of the above-described embodiment.

[0088] The program module 1093 and program data 1094 may not necessarily be stored in the hard disk drive 1090, but may also be stored in, for example, a removable storage medium and read by the CPU 1020 via the disk drive 1100 or the like. Alternatively, the program module 1093 and program data 1094 may be stored in another computer connected via a network (such as a local area network (LAN) or a wide area network (WAN)). The program module 1093 and program data 1094 may then be read by the CPU 1020 from the other computer via the network interface 1070.

[0089] 10 Information providing device 111, 112 Input unit 113, 114 Output unit 121 Scope generating unit 122 Scope setting unit 123 Information collecting unit 124 Answer determining unit 125 Determination result providing unit 131 Manual DB 132 User DB

Claims

1. An information providing device comprising: a scope generation unit that generates multiple scopes related to a problem based on an inquiry message related to the problem; an information collection unit that generates a question inquiring about the situation of the problem and a user targeted by the question for at least one of the multiple scopes, sends the question to the user, and obtains an answer to the question; and an answer determination unit that determines an action to take to solve the problem based on the answer.

2. The information providing device according to claim 1, characterized in that the scope generation unit generates multiple scopes related to the problem by inputting a prompt to the LLM instructing the LLM to generate multiple scopes with different sizes of impact ranges of the problem.

3. The information providing device according to claim 1, characterized in that the answer determination unit determines that the action is one of resetting the scope by the scope generation unit, acquiring further answers by the information collection unit, and performing termination processing.

4. The information providing device according to claim 1, further comprising a judgment result providing unit that creates a summary of the answers acquired by said information collecting unit and outputs the created summary.

5. An information provision method executed by an information provision device, comprising: a scope generation process for generating multiple scopes related to a problem based on an inquiry message related to the problem; an information collection process for generating a question inquiring about the situation of the problem and a user targeted by the question for at least one of the multiple scopes, sending the question to the user, and obtaining an answer to the question; and an answer determination process for determining an action to be taken to solve the problem based on the answer.

6. An information provision program that causes a computer to execute the following steps: a scope generation step of generating multiple scopes related to a problem based on an inquiry message about the problem; an information collection step of generating a question inquiring about the situation of the problem and a user targeted by the question for at least one of the multiple scopes, sending the question to the user, and obtaining an answer to the question; and an answer determination step of determining an action to take to solve the problem based on the answer.

Images