Information processing apparatus

Abstract

To improve the accuracy of answers from a large-scale language model. An information processing apparatus includes a search unit that searches a database based on a query sentence input by a user, an input unit that inputs text data that is part of a search result of the database by the search unit and the query sentence into a large-scale language model, an acquisition unit that acquires an answer to the query sentence generated by the large-scale language model, a display unit that displays the answer, and an acceptance unit that can accept an input by the user. In a case where the acceptance unit accepts an input by the user indicating a selection of the displayed answer, the display unit displays a search result.

Description

Technical Field

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

[0002] As such a device, for example, the following device has been proposed: a language model generates document-based query data, and the pairing of documents and query data is used for learning the search model of the chatbot (see Patent Document 1).

[0003] Patent Document 1: Japanese Patent Application Publication No. 2023-076413 Summary of the Invention

[0004] As a chatbot, a mechanism called Retrieval-Augmented Generation (RAG) is proposed: a unique information source is assigned to the Large Language Model (LLM) by combining a large-scale language model (LLM) with a search for a specific information source (hereinafter, appropriately referred to as the "knowledge base"). In chatbots using RAG, a unique information source is assigned to the LLM by inputting a portion of the search results from the knowledge base into the LLM. However, the information the user expects is sometimes not input into the LLM. As a result, the responses generated by the LLM sometimes differ from the user's requested responses. Furthermore, a large-scale language model refers to a language model built using very large datasets and deep learning techniques.

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

[0006] An information processing apparatus according to one aspect of the present invention includes: a search unit that searches a database based on a query input by a user; an input unit that inputs text data, which is part of the search results of the database by the search unit, and the query into a large-scale language model; an acquisition unit that acquires an answer to the query generated by the large-scale language model; a display unit that displays the answer; and a receiving unit that can accept input from the user, wherein if the receiving unit receives input from the user indicating that the user has selected the answer to be displayed, the display unit displays the search results. 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.

[0010] Figure 4 This is a diagram showing another example of an image. Detailed Implementation

[0011] refer to Figures 1 to 4 The implementation methods related to the information processing apparatus will be described. Figure 1 In this system, information processing system 1 includes information processing device 10, server 20, and knowledge base 30. Information processing device 10, server 20, and knowledge base 30 are configured to communicate with each other via network NW. Server 20 is a server used for applying Large Scale Language Model (LLM). Therefore, server 20 can be called an LLM server. Alternatively, server 20 can be a cloud server.

[0012] Multiple text data entries can be registered in knowledge base 30. Each text data entry can be fragmented data generated by segmenting a document. This fragmented data can be called a "block". Furthermore, as specific examples of document segmentation methods, methods that segment by a certain length (in other words, fixed length), methods that segment by sentence based on delimiters, and methods that segment based on structures such as Markdown can be given. Each text data entry can be vectorized text data. That is, knowledge base 30 can be a vector database / vector storage.

[0013] (Information processing device 10)

[0014] exist Figure 1 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 accepting 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 may 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 and 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] Server 20 and knowledge base 30 provide a chatbot service using RAG. To utilize the chatbot service, the processing unit 11 of the information processing device 10 includes a search unit 111, an input unit 112, an acquisition unit 113, and a modification unit 114 (see reference). Figure 2 The search unit 111, input unit 112, acquisition unit 113, and modification unit 114 can be implemented as the aforementioned logic function blocks. Alternatively, at least one of the search unit 111, input unit 112, acquisition unit 113, and modification unit 114 can be implemented as a physical processing circuit. At least one of the search unit 111, input unit 112, acquisition unit 113, and modification unit 114 can be implemented in a manner that combines logic function blocks and physical processing circuits.

[0025] For example, a user can utilize a chatbot service via information processing device 10. In this case, the user can input a query statement via input device 14 of information processing device 10. Here, "query statement" is not limited to interrogative sentences. For example, a "query statement" can be a statement containing requests, instructions, commands, etc., such as "Please tell me about ****" or "Please answer about ****". Therefore, a "query statement" is not limited to interrogative sentences; it includes, for example, the concept of statements containing requests, instructions, commands, etc. That is, a "query statement" can refer to a statement requesting an answer from the other party.

[0026] The search unit 111 of the information processing apparatus 10 can search the knowledge base 30 based on a query statement. For example, the search unit 111 can calculate a search score indicating the degree of relevance between the query statement and text data registered in the knowledge base 30. The search unit 111 can extract text data with a search score of a predetermined value or higher as text data related to the query statement. The search unit 111 can output search results including the extracted text data and its search score. That is, in the search results, text data and its search score can be associated with each other. In addition, the search unit 111 can extract multiple pieces of text data related to the query statement. That is, the search unit 111 can extract more than one piece of text data related to the query statement from the knowledge base 30. Furthermore, the method for calculating the search score can be applied to various existing methods. Therefore, a detailed description of the method for calculating the search score is omitted.

[0027] The input unit 112 of the information processing device 10 sends a prompt, including a query statement and text data as part of the search results, to the server 20 via the communication device 13. As a result, the prompt is input into a large-scale language model. Here, the text data included in the prompt may be a predetermined number of text data with relatively high search scores from the text data included in the search results (e.g., the top 5 text data with the highest search scores). That is, not all text data included in the search results is necessarily included in the prompt.

[0028] Server 20 sends the response to the query statement generated by a large-scale language model to information processing device 10. Acquisition unit 113 of information processing device 10 acquires the response by receiving it from the large-scale language model. Processing unit 11 of information processing device 10 controls display device 151 to display the response. As a result, the response can be displayed on display device 151. Figure 3 Image 210 is shown. For example, the answer generated by a large-scale language model can be displayed in region 211 of image 210.

[0029] However, the answers generated by the large-scale language model sometimes differ from the answers requested by the user. In this case, in this embodiment, when the user selects an answer generated by the large-scale language model included in the image 210 via the input device 14 (e.g., when clicking on region 211), the computing device 11 controls the display device 151 to display an image 220 representing the search results performed by the search unit 111 (see reference). Figure 4 Additionally, in image 220, the text data included in the aforementioned prompt (in other words, the text data input into a large-scale language model) and the text data not included in the prompt can be visually distinguished.

[0030] By displaying the image 220 on the display device 151, the user can see the text data input into the large-scale language model. For example, if the user realizes that the text data they expect has not been input into the large-scale language model, the user can change the search rating associated with each piece of text data included in the search results. Specifically, the user can instruct the search rating to be changed via the input device 14 so that the text data they expect is input into the large-scale language model (in other words, so that the text data they expect is included in the prompt). As a result, the modification unit 114 of the information processing device 10 can change the search rating included in the search results based on the user's instructions.

[0031] If the search score included in the search results is changed, the input unit 112 selects one or more text data items included in the prompt from the text data included in the search results based on the changed search score. The input unit 112 sends the prompt, including the query statement and the selected one or more text data items, to the server 20 via the communication device 13. As a result, the prompt is input into the large-scale language model.

[0032] When the search score is changed, the processing unit 11 can store one or more text data selected based on the changed search score in association with the query statement in the storage unit 12. Then, when the user inputs a new query statement via the input device 14, the search unit 111 can search the knowledge base 30 based on the new query statement. The search unit 111 can also determine whether a query statement similar to the new query statement is stored in the storage unit 12. If it is determined that a query statement similar to the new query statement is stored in the storage unit 12, the search unit 111 can adjust the search score included in the search results of the knowledge base 30 based on the new query statement, based on one or more text data associated with the query statement similar to the new query statement. For example, the search unit 111 can adjust the search score to include the aforementioned associated text data in the prompts.

[0033] (Technical effect)

[0034] In this embodiment, by inputting prompts including one or more text data selected based on search ratings changed by the user into the large-scale language model, the large-scale language model's answer to the user's query is regenerated. In this case, since the text data desired by the user is input into the large-scale language model, it is expected that the regenerated answer will be the answer requested by the user. Furthermore, in this embodiment, one or more text data selected based on the changed search ratings are stored in the storage device 12 in association with the query. Then, when a new query is input, the search ratings included in the search results of the knowledge base 30 based on the new query can be adjusted based on one or more text data associated with query statements similar to the new query. Therefore, it is expected that the answer from the large-scale language model for the new query will be the answer requested by the user. Thus, according to the information processing apparatus 10 according to this embodiment, the answer accuracy of the large-scale language model can be improved.

[0035] Hereinafter, various aspects of the invention derived from the embodiments described above will be described.

[0036] An information processing apparatus according to one aspect of the present invention includes: a search unit that searches a database based on a query input by a user; an input unit that inputs text data, which is part of the search results of the database by the search unit, and the query into a large-scale language model; an acquisition unit that acquires an answer to the query generated by the large-scale language model; a display unit that displays the answer; and a receiving unit that can accept input from the user, wherein if the receiving unit receives input from the user indicating that the user has selected the answer to be displayed, the display unit displays the search results.

[0037] In the above embodiments, "knowledge base 30" is equivalent to "database" in one example, "search unit 111" is equivalent to "search unit" in one example, "input unit 112" is equivalent to "input unit" in one example, "acquisition unit 113" is equivalent to "acquisition unit" in one example, "input device 14" is equivalent to "accepting unit" in one example, and "display device 151" is equivalent to "display unit" in one example.

[0038] In the information processing apparatus described above, the search results may include text data and a search score indicating the degree of relevance between the text data and the query statement.

[0039] The information processing apparatus described above may include a modification unit that modifies the search score when the receiving unit receives input from the user indicating a change in the search score. In the above embodiment, "modification unit 114" is an example of "modification unit".

[0040] In the information processing apparatus described above, when the search score is changed, the input unit inputs one or more text data selected based on the changed search score and the query statement into the large-scale language model.

[0041] The information processing apparatus described above may include a storage unit that associates one or more text data selected based on the modified search score with the query statement. In the above embodiment, "storage device 12" is an example of "storage unit".

[0042] The present 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 apparatuses that accompany such modifications are also included within the technical scope of the present invention.

[0043] Symbol Explanation

[0044] 1-Information processing system, 10-Information processing device, 20-Server, 30-Knowledge base, 111-Search unit, 112-Input unit, 113-Acquisition unit, 114-Change unit.

Claims

1. An information processing device, characterized in that, have: The search unit searches the database based on a query entered by the user; An input unit that inputs text data, which is part of the search results from the search unit to the database, and the query statement into a large-scale language model; The acquisition unit acquires the response to the query statement generated by the large-scale language model; A display unit displays the answer; and The receiving unit is capable of accepting the user's input. When the receiving unit receives input from the user indicating that they have selected the answer to be displayed, the display unit displays the search results.

2. The information processing device according to claim 1, characterized in that, The search results include text data and a search score indicating the degree of relevance between the text data and the query statement.

3. The information processing device according to claim 2, characterized in that, The system includes a modification unit that, when the receiving unit receives input from the user indicating a change to the search rating, modifies the search rating.

4. The information processing apparatus according to claim 3, characterized in that, In the event that the search score is changed, the input unit will input one or more text data and the query statement selected based on the changed search score into the large-scale language model.

5. The information processing apparatus according to claim 4, characterized in that, It has a storage unit that stores one or more text data selected based on the changed search score in association with the query statement.

Images