Information processing system

Abstract

To improve the response accuracy of large-scale language models. [Solution] The information processing system (1) includes an acquisition means (11) for acquiring one or more text data contained in one workspace from among a plurality of workspaces (WS) contained in a database (40), a classification means (12) for classifying one or more text data, and an assignment means (13) for assigning classification information based on the classification results to one workspace.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing systems.

Background Art

[0002] As this type of system, for example, a system has been proposed in which a query data based on a document is generated in a language model, and a pair of the document and the query data is used for learning a search model for a chatbot (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] As a chatbot, a chatbot using a mechanism (Retrieval-Augmented Generation: RAG) that combines a large language model (Large Language Models: LLM) and retrieval of a specific information source (hereinafter, appropriately referred to as "knowledge base") to provide the large language model with a unique information source has been proposed. For example, the user may specify the knowledge base to be searched. In this case, the knowledge base specified by the user may not contain the information necessary for the large language model to generate an answer. As a result, the answer accuracy of the large language model may decrease. Note that a large language model is a language model constructed using a very large dataset and deep learning technology.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide an information processing system capable of improving the answer accuracy of a large language model. [Means for solving the problem]

[0006] An information processing system according to one aspect of the present invention comprises: acquisition means for acquiring one or more text data contained in one workspace among a plurality of workspaces contained in a database; classification means for classifying the one or more text data; and assignment means for assigning classification information based on the results of the classification to the one workspace. [Brief explanation of the drawing]

[0007] [Figure 1] This is a diagram showing the configuration of an information processing system according to an embodiment. [Figure 2] This is a block diagram showing an example of a computer configuration. [Figure 3] This is a flowchart showing the operation of the information processing system according to the embodiment. [Modes for carrying out the invention]

[0008] Embodiments of the information processing system will be described with reference to Figures 1 to 3. In Figure 1, the information processing system 1 comprises an information processing device 10, an information processing device 20, a server 30, and a database 40. The information processing device 10, the information processing device 20, the server 30, and the database 40 are configured to communicate with each other via a network NW.

[0009] Server 30 is a server for running a Large-Scale Language Model (LLM). Therefore, Server 30 may be referred to as an LLM server. Furthermore, Server 30 may be a cloud server.

[0010] Database 40 contains multiple workspaces WS. Database 40 may be implemented by a single device (e.g., a server) or by multiple devices. Each of the multiple workspaces WS may contain data registered by a designated person (e.g., a person with access rights to a designated workspace).

[0011] For example, when a designated person registers a document as data in a designated workspace (e.g., workspace WSx), the document may be divided into multiple fragmented data. These fragmented data may be referred to as "chunks." Methods for dividing a document include, for example, dividing it at a fixed length, dividing it at the sentence level based on sentence delimiters, or dividing it based on structure such as Markdown. Each of the multiple fragmented data may be vectorized and registered in the database 40. In other words, the database 40 may be a vector database / vector store.

[0012] Server 30 and database 40 provide a chatbot using RAG. Database 40 may also be referred to as a knowledge base.

[0013] At least one of the information processing devices 10 and 20 may be implemented by a computer COM as shown in Figure 2. The computer COM comprises an arithmetic unit 110, a storage device 120, a communication device 130, an input device 140, and an output device 150. The arithmetic unit 110, storage device 120, communication device 130, input device 140, and output device 150 are connected via a data bus 160. The computer COM may be a personal computer, a tablet terminal, or a smartphone.

[0014] The arithmetic unit 110 may have a processor. The arithmetic unit 110 may have a single processor or multiple processors. In other words, the arithmetic unit 110 may have one or more processors. Furthermore, the processor may be a multi-core processor. If the arithmetic unit 110 has a single processor that is a multi-core processor, then logically, the arithmetic unit 110 can be said to have multiple processors.

[0015] The processor may be at least one of the following: CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), and TPU (Tensor Processing Unit).

[0016] The storage device 120 may be at least one of the following: RAM (Random Access Memory), ROM (Read Only Memory), hard disk drive, magneto-optical disk drive, SSD (Solid State Drive), and optical disk array. In other words, the storage device 120 may be implemented by a single device or by multiple devices.

[0017] The communication device 130 may be capable of communicating with devices outside of the computer COM. Furthermore, the communication device 130 may use either wired or wireless communication.

[0018] The input device 140 is a device capable of receiving information input to the computer COM from an external source. The input device 140 may include an operating device that can be operated by the user of the computer COM (e.g., a keyboard, mouse, touch panel, etc.). The input device 140 may include a recording medium reader capable of reading information recorded on a recording medium that can be attached to and detached from the computer COM, such as a USB (Universal Serial Bus) memory. Furthermore, when information is input to the computer COM via the communication device 130 (in other words, when the computer COM acquires information via the communication device 130), the communication device 130 may function as an input device.

[0019] The output device 150 is a device capable of outputting information to the outside of computer COM. The output device 150 may have a display device capable of outputting visual information such as characters and images as the above information. The output device 150 may also have a speaker capable of outputting auditory information such as sound as the above information. The output device 150 may also have a vibration motor capable of outputting tactile information such as vibration as the above information. The output device 150 may also have a printer. The output device 150 may be capable of outputting information to a recording medium that can be attached to and detached from computer COM, such as a USB memory stick. When computer COM outputs information via communication device 130, communication device 130 may function as an output device.

[0020] The storage device 120 is capable of storing desired data. The storage device 120 may store the computer program CP that the arithmetic unit 110 will execute. The storage device 120 may temporarily store data that the arithmetic unit 110 will use temporarily when the arithmetic unit 110 is executing the computer program CP.

[0021] Further, the computer program CP may be recorded on a computer-readable and non-transitory recording medium. In this case, the recording medium may be read using a recording medium reading device (not shown) provided in the computer COM, and the computer program CP may be stored in the storage device 120. As the recording medium, at least one of an optical disk, a magnetic medium, a magneto-optical disk, a semiconductor memory, and any other medium capable of storing a program may be used. Further, the computer program CP may be acquired from a device (not shown) outside the computer COM via the communication device 130. In other words, the computer program CP may be downloaded from an external device to the storage device 120 of the computer COM.

[0022] The arithmetic unit 110 (e.g., a processor) may execute the processing to be performed by the computer COM together with the storage device 120 in which the computer program CP is stored (in other words, together with the storage device 120 and the computer program CP stored in the storage device 120). For example, by executing the computer program CP, the arithmetic unit 110 may realize a logical functional block for executing the processing to be performed by the computer COM (e.g., inside the processor).

[0023] (Operation of the information processing apparatus 10) The information processing apparatus 10 includes an acquisition unit 11, a classification unit 12, and an assignment unit 13. The acquisition unit 11, the classification unit 12, and the assignment unit 13 may be realized as the above-described logical functional blocks. Further, at least one of the acquisition unit 11, the classification unit 12, and the assignment unit 13 may be realized as a physical processing circuit. At least one of the acquisition unit 11, the classification unit 12, and the assignment unit 13 may be realized in a form in which logical functional blocks and physical processing circuits are mixed.

[0024] The acquisition unit 11 of the information processing device 10 acquires one or more data (e.g., text data) contained in one workspace (e.g., workspace WSx) from among multiple workspaces WS contained in the database 40. The classification unit 12 of the information processing device 10 classifies the one or more data acquired by the acquisition unit 11. Various existing methods can be applied to the data classification method. Therefore, a detailed explanation of the data classification method is omitted. The classification unit 12 may classify the data using a learning model constructed by machine learning or the like. The assignment unit 13 of the information processing device 10 assigns classification information based on the classification result by the classification unit 12 to the workspace. For example, the assignment unit 13 may assign the classification information to the workspace as metadata.

[0025] (Operation of the information processing device 20) The information processing device 20 includes a search unit 21 and an input unit 22. The search unit 21 and the input unit 22 may be implemented as the logical functional blocks described above. At least one of the search unit 21 and the input unit 22 may be implemented as a physical processing circuit. At least one of the search unit 21 and the input unit 22 may be implemented in a manner that combines logical functional blocks and physical processing circuits.

[0026] The information processing device 20 is configured to enable the use of chatbots provided by the server 30 and the database 40. Here, each of the multiple workspaces WS included in the database 40 is assigned classification information by the information processing device 10.

[0027] The operation of the information processing device 20 will be explained with reference to the flowchart in Figure 3. The information processing device 20 obtains the question from user U. Here, "question" is not limited to interrogative sentences. For example, "question" may be a sentence that includes expressions such as requests, instructions, or commands, such as "Tell me about ****" or "Answer me about ****". Therefore, "question" is a concept that includes not only sentences in the form of interrogative sentences, but also sentences that include expressions such as requests, instructions, or commands. In other words, "question" may mean a sentence that seeks an answer from the other party.

[0028] When user U enters a question, user U may specify a workspace (for example, workspace WSx) that the search unit 21 of the information processing device 20 should search. When a question is obtained (in other words, when user U enters a question), the search unit 21 determines whether user U has specified a workspace (step S101). If it is determined in step S101 that user U has specified a workspace (step S101: Yes), the search unit 21 searches the specified workspace as a knowledge base based on the question (step S102).

[0029] After processing in step S102, the search unit 21 determines whether the specified workspace contains data related to the question (e.g., text data) (step S103). For example, if data with a search score of a predetermined value or higher is found, the search unit 21 may determine that the specified workspace contains data related to the question. If, in the processing of step S103, it is determined that the specified workspace contains data related to the question (step S103: Yes), the input unit 22 of the information processing device 20 may input the prompt to the large-scale language model by sending a prompt to the server 20 that includes the question and one or more pieces of data related to the question (e.g., text data) found in the specified workspace.

[0030] If, in the process of step S101, it is determined that user U has not specified a workspace (step S101: No), or if, in the process of step S103, it is determined that the specified workspace does not contain data related to the question (step S103: No), the search unit 21 determines the workspace to be searched based on the question and the classification information assigned to each of the multiple workspaces WS (step S104).

[0031] Subsequently, the search unit 21 searches the workspace determined in step S104 based on the question (step S105). The input unit 22 may input the prompt to the large-scale language model by sending a prompt to the server 20 that includes the question and one or more pieces of data related to the question, retrieved from the determined workspace.

[0032] (Technical effects) A chatbot user (e.g., User U) may specify a workspace to search in response to their question. However, the workspace specified by the user may not contain the data necessary for the large-scale language model to generate an answer. Alternatively, the user may not specify a workspace. In these cases, the accuracy of the large-scale language model's answers may decrease.

[0033] In contrast, in this embodiment, the information processing device 10 assigns classification information to each of the multiple workspaces WS contained in the database 40, which corresponds to the knowledge base. If the workspace specified by the user does not contain the data necessary for the large-scale language model to generate an answer, or if the user does not specify a workspace, the search unit 21 of the information processing device 20 determines the workspace to search based on the classification information assigned to each of the multiple workspaces WS. Therefore, in this embodiment, it can be expected that appropriate data will be retrieved from the database 40. Accordingly, the information processing system 1 according to this embodiment can improve the answer accuracy of the large-scale language model.

[0034] Furthermore, each of the multiple workspaces WS in the database 40 may be assigned other classification information, in addition to the classification information assigned by the information processing device 10, that indicates the classification determined by the aforementioned designated person.

[0035] Various aspects of the invention derived from the embodiments described above are described below.

[0036] An information processing system according to one aspect of the invention comprises: an acquisition means for acquiring one or more text data contained in one workspace among a plurality of workspaces contained in a database; a classification means for classifying the one or more text data; and an assignment means for assigning classification information based on the results of the classification to the one workspace. In the above-described embodiment, the "acquisition unit 11" corresponds to an example of the "acquisition means," the "classification unit 12" corresponds to an example of the "classification means," and the "assignment unit 13" corresponds to an example of the "assignment means."

[0037] The information processing system according to the above embodiment may include a search means for searching one or more text data related to a question entered by a user from the database, and an input means for inputting the question and the one or more text data found into a large-scale language model. If the user does not specify a workspace for the search means to search, the search means may determine a workspace to search based on the classification information. In the above embodiment, the "search unit 21" corresponds to an example of the "search means," and the "input unit 22" corresponds to an example of the "input means."

[0038] The information processing system according to the above embodiment may include a search means for searching a database for one or more text data related to a question entered by a user, and an input means for inputting the question and the one or more text data found into a large-scale language model. If the search means searches a workspace specified by the user and the specified workspace does not contain any text data related to the question, the search means may determine a workspace to search based on the classification information.

[0039] The present invention is not limited to the embodiments described above, and can be modified as appropriate without contradicting the gist or idea of ​​the invention as can be read from the claims and specification as a whole. Information processing systems involving such modifications are also included within the technical scope of the present invention. [Explanation of Symbols]

[0040] 1... Information processing system, 10, 20... Information processing device, 11... Acquisition unit, 12... Classification unit, 13... Assignment unit, 21... Search unit, 22... Input unit, 30... Server, 40... Database

Claims

1. A means for retrieving one or more text data items contained in one workspace among multiple workspaces contained in a database, A classification means for classifying one or more text data, A means for assigning classification information based on the results of the classification to the first workspace, An information processing system equipped with the following features.

2. A search means that searches the database for one or more text data related to a question entered by a user, An input means for inputting the aforementioned question and one or more retrieved text data into a large-scale language model, Equipped with, If the user does not specify a workspace to be searched by the search means, the search means determines the workspace to be searched based on the classification information. The information processing system according to claim 1.

3. A search means that searches the database for one or more text data related to a question entered by a user, An input means for inputting the aforementioned question and one or more retrieved text data into a large-scale language model, Equipped with, If the search means searches for a workspace specified by the user, and the specified workspace does not contain text data related to the question, the search means determines the workspace to search based on the classification information. The information processing system according to claim 1.

Images