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

Abstract

To enable a user unsatisfied with an answer to a question to directly inquire of a person familiar with a specialized area relating to contents of the question by presenting, to the user, an inquiry destination for the question or the answer to the question together with the answer corresponding to the question in a chatbot system.SOLUTION: An information processing apparatus 10 for implementing a chatbot system for use in an object organization comprises a document search unit 106 and an inquiry destination search unit 109. The document search unit 106 searches for a document corresponding to an object question from an intra-organization document database. An inquiry destination output unit 111 outputs an intra-organization search inquiry destination group which consists of inquiry destinations to be output together with an answer corresponding to the object question and corresponds to the object question.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to an information processing apparatus, an information processing method, and an information processing program.

Background Art

[0002] An AI (Artificial Intelligence) chatbot system is a system in which the system answers questions from users instead of human experts. Here, the "AI chatbot system" means a system that answers users' questions through intelligent processing using a computer, and it may be expressed as a "chatbot system" without adding "AI". The answers of the AI chatbot system are not always appropriate content that follows the users' questions. Therefore, in an AI chatbot system, in case the users' questions are not solved by the answers of the AI chatbot system, it is often possible to inquire a human operator (inquiry destination). In this case, the inquiry destination is indicated by some method. In the conventional AI chatbot system according to the prior art, the inquiry destination is common through the system. As a specific example, Patent Document 1 discloses an invention for efficiently linking an AI chatbot system and a manned chat, but in Patent Document 1, the operator is single.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In conventional AI chatbot systems, inquiries were directed to a single point of contact. Therefore, when building a complex AI chatbot system based on knowledge spanning multiple specialized fields, users who were dissatisfied with the chatbot's response could not always directly contact an individual with expertise in that specific field. Instead, they had no choice but to first contact a common primary contact person within the system. Consequently, conventional technology required the primary contact person to either attempt to answer questions outside their area of ​​expertise, or to request a secondary contact person with specialized knowledge related to the inquiry.

[0005] Regarding conventional technology, as a concrete example, let's consider the construction of an AI chatbot system that spans multiple related specialized areas (internal company rules and system usage methods), such as rules for travel expense reimbursement and methods for using a travel expense reimbursement system. This example will be explained using Figures 1 and 2.

[0006] In Figure 1, the user first inquired whether telephone charges incurred during a business trip could be reimbursed as travel expenses. This was an inquiry regarding company regulations, and the AI ​​chatbot system appropriately responded that the telephone charges could be reimbursed. The "You" column shows the question entered by the user, and the "Chatbot" column shows the response output based on the question. Next, the user inquired whether personal phone calls could also be included in travel expense reimbursement. However, the AI ​​chatbot system was unable to provide a proper answer and displayed the message, "I'm sorry, but I cannot answer that."

[0007] In Figure 2, after the user went through a question-and-answer session regarding the reimbursement of travel expenses for phone calls during a business trip, the user inquired about how to reimburse the phone calls as travel expenses. The AI ​​chatbot system then responded that the user should select "Other Expenses" using the travel expense reimbursement system. However, the AI ​​chatbot system was unable to properly answer which screen (of the travel expense reimbursement system) to select "Other Expenses," and instead displayed "I'm sorry, but I cannot answer that."

[0008] In both the example shown in Figure 1 and the example shown in Figure 2, the AI ​​chatbot system's response fails to resolve the user's problem. Therefore, the user wants to resolve the issue by contacting a human. In the example shown in Figure 1, the user's problem concerns company regulations regarding whether personal phone calls are eligible for travel expense reimbursement; therefore, it is desirable that the inquiry be handled by someone familiar with the company regulations. On the other hand, in the example shown in Figure 2, the user's problem concerns screen operation in the travel expense reimbursement system; therefore, it is desirable that the inquiry be handled by someone familiar with how to use the travel expense reimbursement system, rather than with company regulations. However, conventional AI chatbot systems cannot change the contact point depending on the question and answer, so they are forced to use a single contact point (the General Affairs Department in the examples shown in Figures 1 and 2) to handle inquiries. Therefore, depending on the area of ​​expertise related to the inquiry (such as company regulations or system operation methods), the person at the contact point may not be able to provide a sufficient answer. Thus, when an AI chatbot system spanning multiple specialized fields has a single point of contact, users had to either research the contact points themselves (through means other than using the chatbot) and use the appropriate contact point themselves, or the AI ​​chatbot system's designated single contact point had to select a specialist for the relevant field, allowing the user to contact the selected specialist or introduce the user to that specialist.

[0009] This disclosure aims to enable users to directly contact experts in the relevant field if they are not satisfied with the answer provided in the chatbot system, by presenting them with a contact point for inquiries related to the question or the answer, along with the answer to the question. [Means for solving the problem]

[0010] The information processing device related to this disclosure is An information processing device for realizing a chatbot used in a target organization, A document search unit searches for one or more documents corresponding to the target question input to the information processing device from an internal document database in which one or more internal documents relating to the target organization are registered, as an internal search document group. Based on the aforementioned internal search document set, an answer processing unit generates an answer corresponding to the target question as an internal answer, A query search unit searches for one or more query destinations, which are output together with the internal response based on the internal search document group, in an internal query destination database where query destinations corresponding to each document registered in the internal document database are registered, as an internal search query destination group corresponding to the target question. The aforementioned A query destination output unit that outputs a group of query destinations for internal organization searches. An information processing device equipped with the following features. [Effects of the Invention]

[0011] According to this disclosure, the information processing device for realizing a chatbot includes an inquiry destination output unit that outputs one or more inquiry destinations corresponding to the target question, which are output along with the internal response. Accordingly, according to this disclosure, a chatbot system can provide users with a contact point for inquiries related to the question or the answer to the question, along with an answer to the question, so that if a user is not satisfied with the answer, they can directly contact someone who is knowledgeable in the area of ​​expertise related to the content of the question. [Brief explanation of the drawing]

[0012] [Figure 1] Figure for explaining problems related to the prior art. [Figure 2] Figure for explaining problems related to the prior art. [Figure 3] Figure showing a configuration example of the chatbot system 90 according to Embodiment 1. [Figure 4] Figure for explaining the inquiry destination related information 3 according to Embodiment 1. [Figure 5] Figure showing a hardware configuration example of the information processing apparatus 10 according to Embodiment 1. [Figure 6] Flowchart showing the operation of the information processing apparatus 10 according to Embodiment 1. [Figure 7] Figure for explaining the effects of Embodiment 1. [Figure 8] Figure for explaining the effects of Embodiment 1. [Figure 9] Figure for explaining the effects of Embodiment 1. [Figure 10] Figure showing a hardware configuration example of the information processing apparatus 10 according to a modification example of Embodiment 1. [Figure 11] Figure showing a configuration example of the chatbot system 90 according to Embodiment 2. [Figure 12] Figure for explaining the effects of Embodiment 2.

Modes for Carrying Out the Invention

[0013] In the description of the embodiments and the drawings, the same elements and corresponding elements are denoted by the same reference numerals. The description of the elements denoted by the same reference numerals may be omitted or simplified as appropriate. The arrows in the figures mainly indicate the flow of data or the flow of processing. Also, "section" may be appropriately read as "circuit", "step", "procedure", "process", or "circuitry".

[0014] s Embodiment 1. Hereinafter, this embodiment will be described in detail with reference to the drawings.

[0015] ***Description of the Configuration*** Figure 3 shows an example configuration of the chatbot system 90 according to this embodiment. The chatbot system 90 comprises an information processing device 10, an inquiry application 11, and a document capture tool 12. The chatbot system 90 is also called an AI (Artificial Intelligence) chatbot system. Two or more elements constituting the chatbot system 90 may be configured integrally.

[0016] The inquiry application 11 is an application that receives a question from user 1 and uses the information processing device 10 to present user 1 with an answer corresponding to the content of the question received from user 1, along with the document from which the answer was based. The document from which the answer was based may be at least a part of the actual document, or it may be source information. The inquiry application 11 may accept questions via text input or via voice input. The answer may be in the form of text information, information including images, or voice information.

[0017] The document import tool 12 is a tool that imports the document group 2 and the contact information related information 3 into the information processing device 10. Document Group 2 consists of internal organizational documents and comprises one or more documents that can serve as the basis for a response. An organization can be, for example, a company, a department, an educational institution, or a government agency, but it can also be the entirety of a service provider and users related to a service provided via the internet. Internal organizational documents are documents held within the organization, and specific examples include company regulations and information regarding internal systems. Contact Information 3 consists of information related to the contact person corresponding to each document included in Document Group 2. In particular, Contact Information 3 includes related information concerning internal company information. The term "document" is not limited to text information; it can also refer to information such as web pages, audio files, or video files.

[0018] The information processing device 10 comprises a document registration unit 101, a query registration unit 103, a question processing unit 105, a document search unit 106, a response processing unit 108, a query search unit 109, and a query output unit 111. The information processing device 10 also stores a document database 102 and a query-related database 104. DB is an abbreviation for database. The information processing device 10 may use a generation AI such as ChatGPT, or it may use other means. The information processing device 10 is a device for realizing a chatbot used in the target organization.

[0019] The document registration unit 101 imports the documents entered from the document import tool 12 into the document database 102.

[0020] Document DB102, also known as the internal organization document database, consists of documents that the AI ​​chatbot refers to when answering questions. The internal organization document database contains one or more internal organization documents related to the target organization.

[0021] The contact registration unit 103 registers the contacts for each document registered in the document database 102 in the contact information database 104, based on the contact information related information 3, in conjunction with the document registration unit 101 registering the documents input from the document import tool 12 in the document database 102. Note that there may be multiple contacts for each document being registered. The Inquiry Contact Database (DB104) is also called the Internal Inquiry Contact Database. The Internal Inquiry Contact Database contains the contact information corresponding to each document registered in the Internal Document Database.

[0022] Figure 4 shows a specific example of contact information related to inquiry 3. In this example, contact information 3 exists independently of the document data registered when each document is registered, and the contact registration unit 103 registers the contact corresponding to each document in the contact information database 104. However, there are other methods for associating documents with contacts. As a concrete example, if contact information is included in each document (for example, by assigning a specific tag), when a document is selected as the source of a response, the contact information corresponding to that document can be obtained from the information contained in that document (for example, by extracting it from the document using the assigned tag as a clue). In other words, contact-related information 3 may be included in document group 2. The key point of this embodiment is not the method of associating documents with contact information, but rather the setting of contact information appropriate to each document. Therefore, the method of associating documents with contact information is not essential to this embodiment.

[0023] The question processing unit 105 generates a search query (search expression) corresponding to the question entered from the inquiry application 11.

[0024] The document search unit 106 searches the document database 102 for one or more documents corresponding to the target query entered into the information processing device 10, as an internal search document group. As a specific example, the document search unit 106 is a document search engine that searches the document database 102 based on a search query received from the query processing unit 105 and obtains search results 107.

[0025] Result 107 is data that shows one or more documents that form the basis of the answer corresponding to the question.

[0026] The response processing unit 108 generates an answer corresponding to the target question as an internal response, based on the internal search document set. As a specific example, the answer processing unit 108 processes the search results 107 to generate an answer corresponding to the question. The answer is a summary of each document shown in search result 107 as an example.

[0027] The inquiry search unit 109 searches the inquiry-related DB 104 for one or more inquiry contacts output along with the internal answer based on the target question, as an internal search inquiry contact group. The inquiry search unit 109 may also search for the internal search inquiry contact group based on the internal search document group. The internal search inquiry contact group corresponds to the target question. The query search unit 109 is a query search engine that, after the document search unit 106 obtains the search result 107, searches for the query contact 110 from the query contact related DB 104 for each document included in the search result 107. The query search unit 109 may also search for the query contact 110 based on a question. Searching for the query contact 110 based on the search result 107 is equivalent to searching for the query contact 110 based on the question corresponding to the search result 107.

[0028] Contact number 110 is the contact number found by the contact number search unit 109, and is output along with the answer corresponding to the question.

[0029] The query destination output unit 111 outputs a group of internal search query destinations. The group of internal search query destinations may consist of query destinations obtained without constructing a physically independent DB (query destination DB 104), such as by searching from the document itself as described in paragraph

[0022] . The query destination output unit 111 outputs, as a specific example, the group of internal search query destinations found by the query destination search unit 109, i.e., the query destination 110.

[0030] In addition to processing questions and answers as described above, AI chatbot systems may also have mechanisms to record a history of consecutive questions and reflect that history in the answer to the next question, or mechanisms to learn based on questions and answers. Such mechanisms can be added to this embodiment, but a description of such mechanisms will be omitted in this embodiment.

[0031] Figure 5 shows an example of the hardware configuration of the information processing device 10 according to this embodiment. The information processing device 10 consists of a computer. The information processing device 10 may consist of multiple computers. Note that the document DB 102 and the query-related DB 104 do not necessarily have to reside in the same computer as the computer that implements the information processing device 10.

[0032] As shown in this figure, the information processing device 10 is a computer equipped with hardware such as a processor 51, memory 52, auxiliary storage device 53, input / output interface 54, and communication device 55. These hardware components are connected as appropriate via signal lines 59.

[0033] The processor 51 is an integrated circuit (IC) that performs arithmetic operations and controls the hardware of the computer. Specific examples of the processor 51 include a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or a GPU (Graphics Processing Unit). The information processing device 10 may include multiple processors that replace the processor 51. The multiple processors share the role of the processor 51.

[0034] Memory 52 is typically a volatile storage device, specifically RAM (Random Access Memory). Memory 52 is also called main memory. Data stored in memory 52 is saved to auxiliary storage device 53 as needed.

[0035] The auxiliary storage device 53 is typically a non-volatile storage device, specifically a ROM (Read Only Memory), an HDD (Hard Disk Drive), or flash memory. Data stored in the auxiliary storage device 53 is loaded into memory 52 as needed. The memory 52 and the auxiliary storage device 53 may be configured as a single unit.

[0036] Input / Output IF54 is a port to which input and output devices are connected. A specific example of an input / output IF54 is a USB (Universal Serial Bus) terminal. Specific examples of input devices include a keyboard and mouse. Specific examples of output devices include a display.

[0037] The communication device 55 is a receiver and a transmitter. Specific examples of the communication device 55 include a communication chip or a NIC (Network Interface Card).

[0038] Each part of the information processing device 10 may use the input / output IF 54 and the communication device 55 as appropriate when communicating with other devices.

[0039] The auxiliary storage device 53 stores the information processing program. The information processing program is a program that enables the computer to implement the functions of each part of the information processing device 10. The information processing program is loaded into memory 52 and executed by the processor 51. The functions of each part of the information processing device 10 are implemented by software.

[0040] Data used when executing an information processing program, and data obtained by executing an information processing program, are appropriately stored in a memory device. Each part of the information processing device 10 utilizes the memory device as appropriate. The memory device, specifically, consists of at least one of the following: memory 52, auxiliary memory device 53, registers in the processor 51, and cache memory in the processor 51. Note that the terms "data" and "information" are sometimes synonymous. The memory device may be independent of the computer. The functions of memory 52 and auxiliary storage device 53 may be implemented by other storage devices.

[0041] The information processing program may be recorded on a computer-readable non-volatile recording medium. Specific examples of non-volatile recording media include optical discs or flash memory. The information processing program may also be provided as a program product.

[0042] ***Explanation of operation*** The operating procedure of the information processing device 10 corresponds to the information processing method. Furthermore, the program that implements the operation of the information processing device 10 corresponds to the information processing program.

[0043] The document registration unit 101 imports the document group 2 into the document database 102 using the document import tool 12 as preparation for answering questions. At the same time, the document registration unit 101 also performs pre-processing so that the question processing unit 105 can appropriately search for documents. This pre-processing varies depending on the technology used in the chatbot system 90, such as the generation AI. However, pre-processing is not essential to this embodiment.

[0044] As preparation for answering questions, the Inquiry Contact Registration Unit 103, in conjunction with the import of document group 2 by the document registration unit 101, uses the document import tool 12 to store the inquiry contacts corresponding to each document imported by the document registration unit 101 in the Inquiry Contact Related DB 104, based on the inquiry contact related information 3.

[0045] Figure 6 is a flowchart illustrating an example of the operation of the information processing device 10. This operation will be explained using Figure 6. After the preliminary preparations for answering questions are complete, when User 1 enters a question into the inquiry application 11, the process shown in this flowchart is executed. As a result, User 1 obtains information indicating the answer to the entered question, the document on which the answer is based, and the contact person for the inquiry corresponding to that answer.

[0046] (Step S101) The question processing unit 105 generates a search query based on the question entered by user 1.

[0047] (Step S102) The document search unit 106 obtains search results 107 by searching the document database 102 using the search query generated by the query processing unit 105. Note that the search results 107 may contain information indicating multiple documents.

[0048] (Step S103) The response processing unit 108 generates a summary sentence to be used as a response based on each document indicated by the search results 107 obtained by the document search unit 106. Here, the response processing unit 108 may generate a summary sentence corresponding to each document indicated by the search results 107, or it may generate a single summary sentence from multiple documents indicated by the search results 107. Here, the mechanism for generating responses varies depending on technologies such as the generation AI used in the chatbot system 90. However, this mechanism is not essential to this embodiment. The generated summary is output to the inquiry application 11.

[0049] (Step S104) The query search unit 109 obtains the query target 110 by searching the query target related DB 104 based on each document indicated by the search results 107 obtained by the document search unit 106. The inquiry destination output unit 111 outputs the inquiry destination 110 obtained by the inquiry destination search unit 109 to the inquiry application 11.

[0050] As described above, user 1 can obtain an answer from the inquiry application 11 that takes into account the search results 107 and the contact information 110 as the answer to their question.

[0051] ***Explanation of the effects of Embodiment 1*** The chatbot system 90 according to this embodiment has the function of changing the contact point according to the user's question and the answer output by the AI ​​chatbot. Therefore, according to this embodiment, it is possible to show a direct contact point according to the content of the question and answer. Therefore, if a complex AI chatbot system based on knowledge spanning multiple specialized fields is constructed using this embodiment, when a user needs to contact a human because they were unable to resolve their problem through the AI ​​chatbot system's response, the system can direct them to a human expert in the relevant field corresponding to their question and answer. As a result, the user can obtain an appropriate answer from the contact person and avoid having to make further inquiries.

[0052] This will be explained in detail using the screen examples in Figures 7 and 8. In Figure 7, the user first inquired whether telephone charges incurred during a business trip could be reimbursed as travel expenses, and received a response from the chatbot system 90 indicating that it could be reimbursed. Subsequently, the user inquired whether personal telephone calls could also be reimbursed as travel expenses, but the chatbot system 90 was unable to provide an appropriate answer. At this point, the chatbot system 90, in its previous response, indicated that the inquiry should be directed to the General Affairs Department. Therefore, the user can inquire with the General Affairs Department about whether or not personal phone calls can be reimbursed as travel expenses.

[0053] In Figure 8, after a question and answer session regarding whether telephone charges incurred during a business trip could be reimbursed as travel expenses, the user inquired about how to reimburse the telephone charges as travel expenses, and the chatbot system 90 responded that the user should select "Other Expenses" using the travel expense reimbursement system. However, the chatbot system 90 was unable to provide an appropriate answer regarding which screen (of the travel expense reimbursement system) to select "Other Expenses." At this point, the chatbot system 90, in its previous response, indicated that the contact point for inquiries was the Information Systems Department. Therefore, the user can inquire with the Information Systems Department about how to select "Other Settlements."

[0054] Comparing Figures 7 and 8, we see that even though the inquiries are directed to the same chatbot system 90, different contact points are shown depending on the user's question and the answer generated by the chatbot system 90. Therefore, according to this embodiment, the user can directly request assistance from the appropriate contact point. Even with an AI chatbot system that spans multiple specialized fields, the chatbot system 90 indicates the appropriate contact point, eliminating the need for users to research contact points themselves (through means other than using the chatbot system 90) and use each contact point individually, or for a fixed primary contact person to select a secondary contact expert appropriate to the field.

[0055] In both examples in Figure 7 and Figure 8, each contact point is for inquiries regarding the immediately preceding question and the answer (which the chatbot system 90 was able to answer appropriately), and not for inquiries regarding the last question that could not be answered appropriately. In this AI chatbot system, the user and the system continue their dialogue to refine the content of the question. Therefore, if the answer immediately preceding a particular response is the last answer that the AI ​​chatbot system was able to provide appropriately, the contact point for that answer is likely to be the appropriate contact point for the question that the system was unable to answer. Thus, the contact point for the answer immediately preceding a particular response can be considered the optimal contact point that the AI ​​chatbot system can suggest for the final question (the one the AI ​​chatbot could not answer appropriately).

[0056] If the chatbot system 90 can associate a contact point with a question rather than with an answer (answer document), then, as mentioned above, it can select a contact point for the current question rather than the contact point for the previous answer. To select a contact point for current inquiries, it's necessary to create a list similar to an FAQ (Frequently Asked Questions) that anticipates user questions. Therefore, it's necessary to anticipate question patterns in advance, or to limit the number of question patterns to the number of patterns that can be used to create an FAQ. Here, an AI chatbot system utilizing generative AI technology is backed by a vast amount of training data and a vast amount of search target data, and has the advantage of being able to generate answers without having to pre-determine user question patterns in the form of FAQs. However, because an AI chatbot system that leverages this advantage cannot pre-limit all question patterns, it cannot show all possible contact points in a way that selects the "contact point for the current question." Here, we can divide the issue into parts addressed based on FAQs and other parts. The former can be addressed by associating question patterns with contact points, while the latter can be addressed using this embodiment. However, this method of resolution is merely an example of an application of this embodiment.

[0057] Furthermore, this embodiment is also effective when the chatbot system 90 presents multiple answers. Figure 9 shows a specific example of when the chatbot system 90 provides multiple answers to a user's question. If user 1 asks a question that is unclear whether it concerns rules regarding the reimbursement of telephone charges as travel expenses, or whether it concerns how to use the travel expense reimbursement system, the document search unit 106 searches both the company regulations and the travel expense reimbursement system manual, and generates multiple answers based on the documents found. The chatbot system 90 then outputs each answer, the document from which each answer is derived, and the contact information for the corresponding inquiry. As shown in the example in Figure 9, if the chatbot system 90 is unable to answer additional questions from the user—that is, if the previous multiple answers are the final answers from the chatbot system 90—the chatbot system 90 will indicate the appropriate contact point for each answer. Therefore, the user can determine for themselves whether they want to ask about the rules or (although they already know that travel expense reimbursement is possible) how to operate the system, and then choose the appropriate contact point to make the appropriate contact. Thus, even if a user asks only one question, if there are multiple possible answers to that question, this embodiment allows for the presentation of multiple separate contact points for inquiries, each related to the answer the user is seeking. Therefore, this embodiment eliminates the need for the user to research contact points themselves (by means other than the chatbot) and use each contact point accordingly, or for a fixed primary contact person to select a secondary contact expert based on their area of ​​expertise.

[0058] ***Other configurations*** <Example 1> Figure 10 shows an example of the hardware configuration of the information processing device 10 according to this modified example. The information processing device 10 includes a processor 51, a processor 51 and memory 52, a processor 51 and auxiliary storage device 53, or a processing circuit 58 instead of a processor 51, memory 52 and auxiliary storage device 53. The processing circuit 58 is hardware that implements at least some of the components of the information processing device 10. The processing circuit 58 may be dedicated hardware, or it may be a processor that executes a program stored in memory 52.

[0059] If the processing circuit 58 is dedicated hardware, specific examples of the processing circuit 58 include a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof. The information processing device 10 may include multiple processing circuits that replace the processing circuit 58. The multiple processing circuits share the role of the processing circuit 58.

[0060] In the information processing device 10, some functions may be implemented by dedicated hardware, while the remaining functions may be implemented by software or firmware.

[0061] The processing circuit 58 can be implemented, in specific examples, by hardware, software, firmware, or a combination thereof. The processor 51, memory 52, auxiliary storage device 53, and processing circuit 58 are collectively referred to as the "processing circuitry." In other words, the functions of each functional component of the information processing device 10 are realized by the processing circuitry. The information processing device 10 according to other embodiments may also have the same configuration as this modified example.

[0062] Embodiment 2. The following will explain the differences from the previously described embodiment, primarily with reference to the drawings.

[0063] ***Explanation of the structure*** Figure 11 shows an example configuration of the chatbot system 90 according to this embodiment. The information processing device 10 according to this embodiment stores a document DB 302 and a contact information DB 304. Furthermore, similar to the document database 102 and the query-related database 104, the document database 302 and the query-related database 304 do not necessarily need to reside on the same computer as the computer that implements the information processing device 10. Document DB302 is also called the external document database. The external document database contains one or more external documents related to the target organization. The contact information database 304 is also called the external contact database. The external contact database contains contact information corresponding to each document registered in the external document database.

[0064] The document import tool 12 according to this embodiment imports document group 202 in addition to document group 2, and imports contact information related information 203 in addition to contact information related information 3.

[0065] Document group 202 consists of external documents and comprises one or more documents that could serve as the basis for a response. External documents are documents other than internal documents and are documents that can be obtained outside the target organization. Specific examples of external documents include documents held by other departments or companies, documents obtainable on the internet, or documents used in chatbot systems 90 used by organizations different from the target organization. Specific examples of documents obtainable on the internet include documents containing tourism-related information.

[0066] Contact-related information 203, like contact-related information 3, is information that associates a document with a contact person. In particular, contact-related information 203 is related information concerning external information. As a specific example, regarding external information such as information on the internet, even if tabular information does not exist as shown in Figure 4, the correspondence between the external information and the contact can be extracted in the same way as the example shown in Figure 4 by considering the provider of the external information as the contact.

[0067] The document registration unit 101 according to this embodiment has the function of registering the document group 202 in the document DB 302, in addition to the functions of the document registration unit 101 according to Embodiment 1.

[0068] The contact registration unit 103 according to this embodiment has, in addition to the functions of the contact registration unit 103 according to Embodiment 1, the function of registering contact-related information 203 in the contact-related DB 304.

[0069] The document search unit 106 according to this embodiment has, in addition to the functions of the document search unit 106 according to Embodiment 1, the function of searching for one or more documents corresponding to the target query from the document DB 302 as an external search document group. As a specific example, the document search unit 106 obtains search results 107 by searching for documents in document DB 102 and document DB 302 based on the search query.

[0070] The answer processing unit 108 according to this embodiment has, in addition to the functions of the answer processing unit 108 according to Embodiment 1, a function to generate an answer corresponding to the target question as an external answer based on a group of externally searched documents. The answer processing unit 108 may generate internal and external answers using a generation AI.

[0071] The inquiry contact search unit 109 according to this embodiment has, in addition to the functions of the inquiry contact search unit 109 according to Embodiment 1, a function to search the inquiry contact-related DB 304 for one or more inquiry contacts output together with the external answer based on the target question, as an external search inquiry contact group. The inquiry contact search unit 109 may also search the external search inquiry contact group based on the external search document group. As a specific example, the query search unit 109 obtains the query target 110 by searching for the query target in the query target related DB 104 and query target related DB 304 based on the document indicated by the search result 107.

[0072] The query destination output unit 111 in this embodiment outputs a group of external search query destinations.

[0073] ***Explanation of operation*** As preparation for answering questions, the document registration unit 101 imports the document group 202 into the document database 302 using the document import tool 12, in addition to the preparations related to Embodiment 1. Note that, for importing documents from outside the organization, it is not necessary to physically import all documents into the document database 302 in advance; the document search unit 106 may register means for accessing documents from outside the organization when searching for documents in the document database 302. As a preliminary step for answering questions, the Inquiry Contact Registration Unit 103, in addition to the preliminary steps described in Embodiment 1, uses the Document Import Tool 12 to store the Inquiry Contacts corresponding to each document imported by the Document Registration Unit 101 in the Inquiry Contact Related DB 304, based on the Inquiry Contact Related Information 203, in conjunction with the import of the document group 202 by the Document Registration Unit 101. Even when the import of external documents is not the import of physical documents but the registration of means of accessing external documents, the Inquiry Contact Related Information can be obtained as described in paragraph

[0022] .

[0074] The following mainly describes the differences in operation from the information processing device 10 according to Embodiment 1.

[0075] (Step S102) This step is the same as step S102 in Embodiment 1. However, the document search unit 106 searches document DB 302 in addition to document DB 102.

[0076] (Step S104) This step is the same as step S104 in Embodiment 1. However, if the search result 107 indicates an external document, the query search unit 109 searches the query-related database 304 to obtain a query 110 corresponding to the external document.

[0077] ***Explanation of the effects of Embodiment 2*** According to this embodiment, the chatbot system 90 can provide answers not only to questions that should be answered based on internal organizational information, but also to questions that concern external organizational information.

[0078] Figure 12 shows a specific example of a case where User 1 asks a question about an external document to the chatbot system 90 according to this embodiment. User 1 is not only asking about internal organizational information, such as whether travel expenses can be reimbursed even if they take a plane to City O, but is also seeking tourist information (information about okonomiyaki restaurants) that cannot be answered based on internal organizational information. The chatbot system 90 can provide answers based on both internal and external organizational documents, even for the questions shown in Figure 12. The chatbot system 90 may provide answers based only on external organizational documents, or it may provide answers based on both internal and external organizational documents. Furthermore, in this embodiment as well, it is possible to indicate multiple contact points depending on the content of the response.

[0079] ***Other Embodiments*** The aforementioned embodiments can be freely combined, any component of each embodiment can be modified, or any component can be omitted in each embodiment. Furthermore, the embodiments are not limited to those shown in Embodiments 1 and 2, and various modifications can be made as needed. The procedures described using flowcharts and the like may be modified as appropriate.

[0080] The various aspects of this disclosure are summarized below as an appendix.

[0081] (Note 1) An information processing device for realizing a chatbot used in a target organization, A document search unit searches for one or more documents corresponding to the target question input to the information processing device from an internal document database in which one or more internal documents relating to the target organization are registered, as an internal search document group. Based on the aforementioned internal search document set, an answer processing unit generates an answer corresponding to the target question as an internal answer, The following are included: an inquiry destination output unit which outputs one or more inquiry destinations that are output together with the aforementioned internal response, and which outputs a group of internal search inquiry destinations corresponding to the target question; An information processing device equipped with the following features.

[0082] (Note 2) The aforementioned information processing device further, Based on the aforementioned target question, the query search unit searches for one or more query destinations as the group of internal search query destinations from the internal query destination database, which contains query destinations corresponding to each document registered in the internal document database. An information processing device as described in Appendix 1, comprising:

[0083] (Note 3) The aforementioned query destination search unit is an information processing device as described in Appendix 2, which searches the group of internal search query destinations based on the group of internal search documents within the organization.

[0084] (Note 4) The document search unit searches for one or more documents corresponding to the target query from an external document database in which one or more external documents relating to the target organization are registered, as an external search document group. The response processing unit generates an answer corresponding to the target question as an external response based on the external search document group, The aforementioned contact search unit searches for one or more contacts that are output together with the external answer based on the target question, as an external search contact group, from the external contact database in which contacts corresponding to each document registered in the external document database are registered. The aforementioned query destination output unit is an information processing device according to Appendix 2 or 3 that outputs the group of external search query destinations.

[0085] (Note 5) The aforementioned inquiry destination search unit is an information processing device described in Appendix 4 that searches the group of external search destinations based on the group of external search documents.

[0086] (Note 6) The response processing unit is an information processing device according to Appendix 4 or 5 that generates the internal response and the external response using generation AI (Artificial Intelligence). [Explanation of Symbols]

[0087] 1 User, 2,202 Document group, 3,203 Inquiry contact related information, 10 Information processing device, 101 Document registration unit, 102,302 Document DB, 103 Inquiry contact registration unit, 104,304 Inquiry contact related DB, 105 Question processing unit, 106 Document search unit, 107 Search results, 108 Answer processing unit, 109 Inquiry contact search unit, 110 Inquiry contact, 111 Inquiry contact output unit, 11 Inquiry application, 12 Document import tool, 51 Processor, 52 Memory, 53 Auxiliary storage device, 54 Input / Output IF, 55 Communication device, 58 Processing circuit, 59 Signal line, 90 Chatbot system.

Claims

1. An information processing device for realizing a chatbot used in a target organization, A document search unit searches for one or more documents corresponding to the target question input to the information processing device from an internal document database in which one or more internal documents relating to the target organization are registered, as an internal search document group. Based on the aforementioned internal search document set, an answer processing unit generates an answer corresponding to the target question as an internal answer, A query search unit searches for one or more query destinations, which are output together with the internal response based on the internal search document group, as an internal search query destination group corresponding to the target question, from an internal query destination database in which query destinations corresponding to each document registered in the internal document database are registered. A query destination output unit that outputs the aforementioned group of internal search query destinations and An information processing device equipped with the following features.

2. The information processing device according to claim 1, wherein the response processing unit generates the internal response using artificial intelligence (Generative AI).

3. The document search unit searches for one or more documents corresponding to the target query from an external document database in which one or more external documents relating to the target organization are registered, as an external search document group. The response processing unit generates an answer corresponding to the target question as an external response based on the external search document group, The aforementioned contact search unit searches for one or more contacts that are output together with the external response, based on the external search document group, as an external search contact group, from the external contact database in which contacts corresponding to each document registered in the external document database are registered. The information processing apparatus according to claim 1, wherein the query destination output unit outputs the group of external search query destinations.

4. The information processing apparatus according to claim 3, wherein the response processing unit generates the internal response and the external response using Artificial Intelligence (Generative AI).

5. The information processing device according to any one of Claims 1 to 4, wherein the target question is a question entered by a user into an inquiry application.

6. An information processing method executed by an information processing device, which is a computer used to implement a chatbot in a target organization, The information processing device searches for one or more documents corresponding to the target question input to the information processing device, as an internal search document group, from an internal document database in which one or more internal documents relating to the target organization are registered. The information processing device generates an answer corresponding to the target question as an internal answer based on the internal search document group, The information processing device consists of one or more query destinations output together with the internal response based on the internal search document group, and searches for one or more query destinations corresponding to the target question from the internal query destination database, which contains query destinations corresponding to each document registered in the internal document database. The information processing device is an information processing method that outputs the group of search destinations within the organization.

7. An information processing program executed by an information processing device, which is a computer used to implement a chatbot in the target organization, A document search process that searches for one or more documents corresponding to the target question input to the information processing device from an internal document database in which one or more internal documents relating to the target organization are registered, as an internal search document group. Based on the aforementioned internal search document set, the response process generates an answer corresponding to the target question as an internal response, Based on the aforementioned internal search document group, a query search process is performed to search for one or more query destinations, which are output together with the internal answer, as an internal search query destination group corresponding to the target question, from the internal query destination database, which contains query destinations corresponding to each document registered in the internal document database. The query destination output process outputs the aforementioned group of query destinations for internal search within the organization. An information processing program that causes the information processing device to execute the above.

Images