Information processing system

Abstract

To improve the answer accuracy of a large-scale language model. An information processing system includes: an extraction unit that, when a first question sentence related to one of a plurality of processes included in a business process is input, extracts first text data related to the first question sentence from one work area corresponding to the one process of a database including a plurality of work areas corresponding to the plurality of processes, respectively; an input unit that inputs the first question sentence and the first text data into a large-scale language model; and an acquisition unit that acquires a first answer to the first question sentence generated by the large-scale language model.

Description

Technical Field

[0001] This invention relates to the technical field of an information processing system. Background Technology

[0002] As such a system, for example, a system for generating documents as prompts input into Large Language Models (LLM) has been proposed (see Patent Document 1).

[0003] Patent Document 1: Japanese Patent No. 7325152 Summary of the Invention

[0004] For example, a chatbot has been proposed that utilizes a mechanism (Retrieval-Augmented Generation: RAG) to endow a large-scale language model with a unique information source (hereinafter, appropriately referred to as a "knowledge base") by combining retrieval with a specific information source. Such chatbots are sometimes used as business support tools. In enterprises, business processes are typically defined for a particular business. When the aforementioned chatbot is used as a business support tool, without any countermeasures, there is a high probability of generating responses that ignore the business process. That is, the accuracy of the large-scale language model's responses may decrease. Furthermore, a large-scale language model refers to a language model constructed using very large datasets and deep learning techniques.

[0005] The present invention was made in view of the above-mentioned problems, and its objective is to provide an information processing system that can improve the response accuracy of large-scale language models.

[0006] An information processing system according to one aspect of the present invention comprises: an extraction unit that, upon inputting a first question related to one of a plurality of processes included in a business process, extracts first text data related to the first question from a work area corresponding to the first process, which includes a database comprising multiple work areas corresponding to the plurality of processes; an input unit that inputs the first question and the first text data into a large-scale language model; and an acquisition unit that acquires a first answer to the first question generated by the large-scale language model. Attached Figure Description

[0007] Figure 1 This is a diagram showing the structure of the information processing system involved in the implementation method.

[0008] Figure 2 This is a block diagram illustrating an example of the structure of the computing device involved in the implementation.

[0009] Figure 3 This is a diagram representing an example of an image being displayed. Detailed Implementation

[0010] refer to Figures 1 to 3 The implementation methods involved in the information processing system are described. Figure 1 In this system, information processing system 1 includes information processing device 10, server 20, and database 30. The information processing device 10, server 20, and database 30 are configured to communicate with each other via network NW.

[0011] Server 20 and database 30 provide a chatbot using RAG. Server 20 is a server used for applying Large Scale Language Models (LLM). Therefore, server 20 can be called an LLM server. Alternatively, server 20 can be a cloud server.

[0012] Database 30 includes multiple work areas WS. Furthermore, database 30 can be implemented by a single device (e.g., a server) or by multiple devices. In this embodiment, the multiple work areas WS correspond to multiple steps included in the business process. One work area (e.g., work area WS1) within the multiple work areas WS includes text data related to one of the multiple steps.

[0013] For example, the text data included in database 30 (in other words, the registered text data) can be fragmented data generated by segmenting documents. Fragmented data can be called "chunks." Furthermore, methods for segmenting documents include, for example, segmenting with a constant length (in other words, fixed length), segmenting by sentence based on sentence delimiters, and segmenting based on structures such as Markdown. Additionally, multiple fragmented data can be vectorized and registered separately in database 30. That is, database 30 can be a vector database / vector storage. Furthermore, database 30 can be called a knowledge base.

[0014] The information processing device 10 includes a computing unit 11, a storage unit 12, a communication unit 13, an input unit 14, and an output unit 15. The computing unit 11, storage unit 12, communication unit 13, input unit 14, and output unit 15 are connected via a data bus 16. Furthermore, the information processing device 10 can be a personal computer, a tablet terminal, or a smartphone.

[0015] The arithmetic unit 11 may have a processor. Furthermore, the arithmetic unit 11 may have a single processor or multiple processors. That is, the arithmetic unit 11 may have more than one processor. Additionally, the processor may be a multi-core processor. In the case where the arithmetic unit 11 has a single processor that functions as a multi-core processor, it can be said that the arithmetic unit 11 logically has multiple processors.

[0016] The processor may be at least one of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a Field Programmable Gate Array (FPGA), and a Tensor Processing Unit (TPU).

[0017] Storage device 12 may be at least one of random access memory (RAM), read-only memory (ROM), hard disk drive, magneto-optical disk drive, solid-state drive (SSD), and optical disk array. That is, storage device 12 may be implemented by a single device or by multiple devices.

[0018] The communication device 13 can communicate with external devices (e.g., server 20) of the information processing device 10. Furthermore, the communication device 13 can perform wired communication or wireless communication.

[0019] Input device 14 is a device capable of receiving information input from an external source to information processing device 10. Input device 14 may include user-operable devices (e.g., keyboard, mouse, touch panel, etc.) of information processing device 10. Input device 14 may include, for example, a recording medium reading device capable of reading information recorded on a recording medium removable from information processing device 10, such as a Universal Serial Bus (USB) memory. Furthermore, when information is input to information processing device 10 via communication device 13 (in other words, when information processing device 10 obtains information via communication device 13), communication device 13 can function as an input device.

[0020] Output device 15 is a device capable of outputting information to the outside of information processing device 10. Output device 15 includes a display device 151 capable of outputting visual information such as characters or images. Additionally, output device 15 may include a speaker capable of outputting auditory information such as sound. Output device 15 may include a vibration motor capable of outputting tactile information such as vibration. Output device 15 may include a printer. Output device 15 can output information to recording media removable from information processing device 10, such as a USB memory. Furthermore, when information processing device 10 outputs information via communication device 13, communication device 13 can function as an output device.

[0021] Storage device 12 is capable of storing desired data. The computer program CP executed by the arithmetic unit 11 can be stored in storage device 12. When the arithmetic unit 11 executes the computer program CP, storage device 12 can temporarily store data temporarily used by the arithmetic unit 11.

[0022] Furthermore, the computer program CP can be recorded on a computer-readable and non-temporary recording medium. In this case, the computer program CP can be stored in the storage device 12 by reading the recording medium using a recording medium reading device (not shown) included in the information processing device 10. Additionally, at least one of optical discs, magnetic media, magneto-optical discs, semiconductor memory, and any medium capable of storing other programs can be used as the recording medium. Furthermore, the computer program CP can also be obtained from an external (not shown) device outside the information processing device 10 via the communication device 13. In other words, the computer program CP can be downloaded from an external device to the storage device 12 of the information processing device 10.

[0023] The arithmetic unit 11 (e.g., a processor) can perform the processing to be performed by the information processing unit 10 together with the storage device 12 storing the computer program CP (in other words, together with the storage device 12 and the computer program CP stored in the storage device 12). For example, the arithmetic unit 11 can implement logical function blocks for performing the processing to be performed by the information processing unit 10 within the arithmetic unit 11 (e.g., within the processor) by executing the computer program CP.

[0024] The arithmetic unit 11 of the information processing device 10 includes a retrieval unit 111, an input unit 112, an acquisition unit 113, and a generation unit 114 (see reference) for the purpose of using the chatbot provided by the server 20 and the database 30. Figure 2The retrieval unit 111, input unit 112, acquisition unit 113, and generation unit 114 can be implemented as the aforementioned logic function blocks. Alternatively, at least one of the retrieval unit 111, input unit 112, acquisition unit 113, and generation unit 114 can be implemented as a physical processing circuit. At least one of the retrieval unit 111, input unit 112, acquisition unit 113, and generation unit 114 can be implemented in a manner that combines logic function blocks and physical processing circuits.

[0025] For example, user U can use information processing device 10 to utilize a chatbot. For instance, user U can use information processing device 10 to input a question related to a step included in a business process into the chatbot. In this case, user U can input the first question via input device 14 of information processing device 10. At this time, user U can input the first question via input device 14 as a question related to the aforementioned step. For example, display device 151 can display images representing multiple steps included in the business process. User U can input the step related to the first question by selecting one step from the image.

[0026] The term "question sentence" is not limited to interrogative sentences. For example, a "question sentence" can be a sentence that expresses a request, instruction, or command, such as "Please tell me about ****" or "Please answer about ****". Therefore, the concept of a "question sentence" is not limited to interrogative sentences; it includes sentences that express requests, instructions, or commands. In other words, a "question sentence" can refer to a sentence that requests an answer from the other party.

[0027] The retrieval unit 111 of the information processing apparatus 10 retrieves the work area WS1 corresponding to a process from among the multiple work areas WS included in the database 30, based on the first question. For example, the retrieval unit 111 can calculate a retrieval score representing the relevance of the first question to the text data included in the work area WS1. The retrieval unit 111 can extract text data with a retrieval score of a predetermined value or higher as text data related to the first question. Furthermore, the retrieval unit 111 can extract multiple pieces of text data related to the first question. That is, the retrieval unit 111 can extract more than one piece of text data related to the first question from the work area WS1. In addition, the method for calculating the retrieval score can be applied to various existing methods. Therefore, a detailed description of the method for calculating the retrieval score is omitted.

[0028] The input unit 112 of the information processing device 10 sends a prompt, including a first question and text data related to the first question, to the server 20 via the communication device 13. As a result, the prompt is input into a large-scale language model. The server 20 sends a first answer generated by the large-scale language model for the first question to the information processing device 10. The acquisition unit 113 of the information processing device 10 acquires the first answer sent by the server 20. The processing unit 11 of the information processing device 10 can control the display device 151 to display the first answer.

[0029] After the acquisition unit 113 acquires the first answer, the generation unit 114 of the information processing device 10 generates a second question related to the next step of a process based on the first answer. That is, the second question is generated automatically. In addition, the generation unit 114 can generate the second question based on the first answer and the first question.

[0030] For example, the business process could be a business process related to regulatory certification. In this case, the business process could include steps such as "searching for relevant regulations" and "interpreting regulations." For example, the first question mentioned above could be a question related to the "searching for relevant regulations" step. In this case, the first answer mentioned above could include the content of the regulations. The next step after the "searching for relevant regulations" step could be the "interpreting regulations" step. For example, the generation unit 114 could generate a question such as "What is ****?" to inquire about the explanation of words included in the regulations, based on the content of the regulations included in the first answer, as the second question mentioned above.

[0031] After generating the second question, the retrieval unit 111 retrieves the work area WS2 corresponding to the next process in the multiple work areas WS included in the database 30, based on the second question. For example, the retrieval unit 111 can calculate a retrieval score indicating the relevance of the second question to the text data included in work area WS2. The retrieval unit 111 can extract text data with a retrieval score of a predetermined value or higher as text data related to the second question.

[0032] The input unit 112 sends a prompt, including a second question and related text data, to the server 20 via the communication device 13. As a result, the prompt is input into a large-scale language model. The server 20 sends a second answer generated by the large-scale language model for the second question to the information processing device 10. The acquisition unit 113 of the information processing device 10 acquires the second answer sent by the server 20. The processing unit 11 can control the display device 151 to display the second answer.

[0033] For example, it can be displayed on display device 151 Figure 3The image 200 shown may include an area 201 displaying a first question, an area 202 displaying a first answer, an area 203 displaying a second question, and an area 204 displaying a second answer.

[0034] After the second question and second answer are displayed on the display device 151, the user U can change the second question via the input device 14. In this case, the input unit 14 can receive input from the user U for changing the second question. When the user U changes the second question, the retrieval unit 111 retrieves the work area WS2 based on the changed second question. For example, the retrieval unit 111 can calculate a retrieval score indicating the relevance of the changed second question to the text data included in the work area WS2. The retrieval unit 111 can extract text data with a retrieval score of a predetermined value or higher as text data related to the changed second question.

[0035] The input unit 112 sends a prompt, including a modified second question and text data related to the modified second question, to the server 20 via the communication device 13. As a result, the prompt is input into a large-scale language model. The server 20 sends a third answer generated by the large-scale language model for the modified second question to the information processing device 10. The acquisition unit 113 of the information processing device 10 acquires the third answer sent by the server 20. The processing unit 11 can control the display device 151 to display the third answer.

[0036] (Technical effect)

[0037] As described above, the database 30 includes multiple work areas WS corresponding to multiple steps included in the business process. The retrieval unit 111 of the information processing device 10 retrieves the work area WS1 corresponding to a step related to the first question input by the user U. The work area WS1 includes text data related to a step. Therefore, by retrieving the work area WS1 by the retrieval unit 111, text data related to the first question and a step is extracted. The input unit 112 of the information processing device 10 sends a prompt including the first question and text data related to the first question and a step to the server 20 (in other words, the prompt is input into the large-scale language model). Therefore, it is expected that the first answer generated by the large-scale language model will be an answer related to a step. Therefore, in the information processing system 1 according to this embodiment, it is possible to generate an appropriate answer to a question specific to a step included in the business process. As a result, according to the information processing system 1 according to this embodiment, the answer accuracy of the large-scale language model can be improved.

[0038] Hereinafter, various aspects of the invention derived from the embodiments described above will be described.

[0039] One aspect of the invention relates to an information processing system comprising: an extraction unit that, upon inputting a first question related to one of a plurality of processes included in a business process, extracts first text data related to the first question from a workspace corresponding to the first process, comprising a database of multiple workspaces corresponding to the plurality of processes; an input unit that inputs the first question and the first text data into a large-scale language model; and an acquisition unit that acquires a first answer generated by the large-scale language model for the first question. In the above embodiment, "retrieval unit 111" is equivalent to an example of "extraction unit," "input unit 112" is equivalent to an example of "input unit," and "acquisition unit 113" is equivalent to an example of "acquisition unit."

[0040] In the information processing system described above, the extraction unit can extract at least the second text data related to the second question corresponding to the first answer from other work areas in the database corresponding to the next process of the first process. The input unit can input the second question and the second text data into the large-scale language model. The acquisition unit can acquire the second answer generated by the large-scale language model for the second question.

[0041] The information processing system described above may include: a receiving unit for receiving user input; and a display unit, wherein, after the receiving unit displays the second question and the second answer, and then receives user input for changing the second question, the extraction unit extracts third text data related to the changed second question from other work areas.

[0042] The input unit can input the modified second question and the third text data into the large-scale language model. In the above embodiment, "input device 14" is equivalent to an example of "receiving unit", and "display device 151" is equivalent to an example of "display unit".

[0043] This invention is not limited to the embodiments described above, and appropriate modifications can be made without departing from the spirit or concept of the invention as read in its entirety from the claims and description. Information processing systems that accompany such modifications are also included within the technical scope of this invention.

[0044] Symbol Explanation

[0045] 1-Information processing system, 10-Information processing device, 20-Server, 30-Database, 111-Retrieval unit, 112-Input unit, 113-Acquisition unit, 114-Generation unit.

Claims

1. An information processing system, characterized in that, have: The extraction unit, upon inputting a first question related to one of the multiple processes included in the business process, extracts first text data related to the first question from a work area corresponding to the one process in a database that includes multiple work areas corresponding to the multiple processes respectively. The input unit inputs the first question and the first text data into a large-scale language model; and The acquisition unit acquires the first answer to the first question generated by the large-scale language model.

2. The information processing system according to claim 1, characterized in that, The extraction unit extracts, from other work areas in the database corresponding to the next process of the first process, at least some second text data related to the second question corresponding to the first answer. The input unit inputs the second question and the second text data into the large-scale language model. The acquisition unit acquires the second answer to the second question generated by the large-scale language model.

3. The information processing system according to claim 2, characterized in that, have: The receiving unit receives user input; and Display unit, After the receiving unit displays the second question and the second answer through the display unit, and then receives input from the user to change the second question, the extraction unit extracts third text data related to the changed second question from the other work areas. The input unit inputs the modified second question and the third text data into the large-scale language model.

Images