Information processing system, information processing program, and information processing method

The information processing system addresses the challenge of understanding diverse scientific literature by using a large-scale language model to generate keywords and organize documents by theme, enhancing the efficiency of hypothesis construction and preliminary research.

JP2024159418AActive Publication Date: 2024-11-08MEMORY LAB INC

Patent Information

Application Number
JP2023172060
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-04-28
Filing Date
2023-10-03
Publication Date
2024-11-08
Estimated Expiration
2043-10-02

AI Technical Summary

Technical Problem

The fragmentation and sophistication of science and technology make it difficult to efficiently understand papers and patents related to fields outside one's own specialty, hindering effective hypothesis construction and preliminary research in business development.

Method used

An information processing system utilizing a large-scale language model to generate keywords and organize documents by theme, enabling users to visualize relationships and retrieve relevant literature through a user-friendly interface, tailored to different knowledge levels.

Benefits of technology

Facilitates efficient retrieval of relevant information across diverse fields, reducing the time required for hypothesis construction and preliminary research by providing organized and understandable summaries of complex literature.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024159418000001_ABST
    Figure 2024159418000001_ABST
Patent Text Reader

Abstract

To provide an information processing system, method, and program for enabling theses and patents in different fields of specialty to be efficiently understood.SOLUTION: An information processing system 1000 in which a server device 100 and a client device 110 are communicably connected via a network 150 and which is also communicably connected to external systems via a network includes at least one processor. The processor is constituted so as to execute each step below. In a first reception step, first information inputted by a user and pertaining to a subject which a user wants to search for is received. In an acquisition step, a plurality of keywords recalled from the subject are acquired via a large-scale language model by using a prescribed prompt including the inputted first information. In a first display step, second information in which subjects are organized for each of a plurality of literature themes is displayed on the basis of the plurality of keywords and a database that stores information on a plurality of literatures.SELECTED DRAWING: Figure 4
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an information processing system, an information processing program, and an information processing method. [Background technology]

[0002] Patent Document 1 discloses a method for collecting data for project implementation. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2010-262519 A Summary of the Invention [Problem to be solved by the invention]

[0004] Hypothesis building and preliminary research are necessary for business development using science and technology. However, there are a huge number of papers and patents related to a certain science and technology. In addition, as technology becomes more specialized and sophisticated, it is difficult to efficiently understand papers and patents related to technologies in fields that are even slightly different from one's own specialty. [Means for solving the problem]

[0005] According to one aspect of the present invention, there is provided an information processing system. The information processing system includes at least one processor. The processor is configured to execute the following steps: In a first reception step, first information related to an item that the user wishes to search for is received, which is input by the user; In an acquisition step, a plurality of keywords associated with the item are acquired through a large-scale language model using a predetermined prompt including the input first information; In a first display control step, second information in which items are organized by theme of a plurality of documents is displayed based on the plurality of keywords and a database storing information on a plurality of documents.

[0006] According to one aspect of the present invention, a more beneficial information processing system and the like can be provided. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 illustrates an example of a system configuration of an information processing system 1000. [Diagram 2] 2 is a diagram illustrating an example of a hardware configuration of a server device 100. FIG. [Diagram 3] FIG. 2 illustrates an example of a hardware configuration of a client device 110. [Figure 4] 1 is a sequence diagram (part 1) showing an example of information processing in the information processing system 1000. FIG. [Diagram 5] FIG. 5 is a diagram showing an example of a screen 500. [Figure 6] 12 is a sequence diagram (part 2) illustrating an example of information processing in the information processing system 1000. FIG. [Figure 7] FIG. 7 is a diagram showing an example of a screen 700. [Figure 8] FIG. 8 is a diagram showing an example of a screen 800. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings. Various characteristic features shown in the following embodiments can be combined with each other.

[0009] In this specification, the term "unit" may include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In addition, various information is handled in this embodiment, and communication and calculation can be performed on a circuit in the broad sense regardless of whether this information is represented by a high or low signal value as a binary bit collection consisting of 0 or 1, represented by a physical numerical value of a signal value, or represented by quantum superposition.

[0010] In addition, a circuit in the broad sense is a circuit realized by at least appropriately combining a circuit, circuitry, a processor, a memory, etc. In other words, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

[0011] In addition, the program for realizing the software appearing in the embodiments may be implemented in a form that is downloadable from a server device, the program may be executed on a cloud computer, or the program may be stored in a non-volatile or volatile non-transitory storage medium and distributed.

[0012] <Embodiment 1> 1. System Configuration Fig. 1 is a diagram showing an example of a system configuration of an information processing system 1000. As shown in Fig. 1, the information processing system 1000 includes, as a system configuration, a server device 100 and a client device 110. The server device 100 and the client device 110 are communicatively connected via a network 150. The information processing system 1000 is communicatively connected to an information processing system 120 via the network 150. Note that the server device 100 may be realized as a cloud system.

[0013] The server device 100 executes the process related to the first embodiment including the modified example described later. More specifically, the server device 100 controls the client device 110 to display screens such as those shown in Fig. 5, Fig. 7, and Fig. 8 described later in response to a request from the client device 110.

[0014] The client device 110 is a terminal used by a user, and displays screens such as those shown in Fig. 5, Fig. 7, and Fig. 8 described below. For simplicity, only one client device 110 is shown in Fig. 1, but the information processing system 1000 may include multiple client devices 110. However, for simplicity, the following description will be given taking one client device 110 as an example.

[0015] The information processing system 120 is an information processing system other than the information processing system 1000, and provides the function of a large-scale language model. The large-scale language model is a type of deep learning model having hundreds of millions to hundreds of billions of parameters. The large-scale language model learns from a large text corpus and is used to perform the task of natural language understanding. The large-scale language model interprets a sentence given in a prompt and generates an appropriate response in the context. An example of the large-scale language model is Generative Pre-trained Transformer (GPT (registered trademark)), and for example, the information processing system 120 provides the function of the large-scale language model as a function of a cloud service.

[0016] 1, a PC (Personal Computer) is shown as an example of the client device 110, but the client device 110 is not limited to a PC and may be a smartphone, a tablet computer, etc. The client device 110 may be, for example, any device that can display information from the server device 100 and allow the user to input items that he or she wishes to search for.

[0017] Here, the information processing system described in the claims may be composed of multiple devices or may be composed of one device. When the information processing system described in the claims is composed of one device, an example of the device is a server device 100. When the information processing system described in the claims is composed of multiple devices, an example of the multiple devices is the server device 100 and the client device 110, or a cloud system that provides the same functions as the server device 100, etc.

[0018] 2. Hardware Configuration (Hardware configuration of the server device 100) 2 is a diagram showing an example of a hardware configuration of the server device 100. The hardware configuration of the server device 100 includes a control unit 210, a storage unit 220, and a communication unit 230. Each of the components will be further described.

[0019] The control unit 210 is a CPU (Central Processing Unit) or the like, and controls the entire server device 100 . The storage unit 220 is any one of a hard disk drive (HDD), a read only memory (ROM), a random access memory (RAM), and a solid state drive (SSD), or any combination thereof, and stores a program and data used when the control unit 210 executes a process based on the program. The control unit 210 executes a process based on the program stored in the storage unit 220, thereby realizing the functions of the server device 100. The storage unit 220 is an example of a storage medium. Note that in this embodiment, the data used when the control unit 210 executes a process based on the program is described as being stored in the storage unit 220, but the data may be stored in a storage unit of another device that can communicate with the server device 100. That is, the data may be stored in a storage unit of any device as long as the control unit 210 can refer to the data. A literature database is stored in the storage unit 220. Details will be described later. The communication unit 230 is a NIC (Network Interface Card) or the like, which connects the server device 100 to the network 150 and controls communication with other devices.

[0020] (Hardware Configuration of Client Device 110) 3 is a diagram showing an example of a hardware configuration of client device 110. Client device 110 includes, as its hardware configuration, a control unit 310, a storage unit 320, an input unit 330, an output unit 340, and a communication unit 350.

[0021] The control unit 310 is a CPU or the like, and controls the entire client device 110 .

[0022] The storage unit 320 is any one of an HDD, a ROM, a RAM, and an SSD, or any combination thereof, and stores programs, data used when the control unit 310 executes processing based on the programs, and the like. The control unit 310 executes processing based on the programs stored in the storage unit 320, thereby realizing the functions of the client device 110. The storage unit 320 is an example of a storage medium.

[0023] The input unit 330 is a keyboard, a mouse, or the like, and inputs information to the client device 110 based on an input operation by a user. The input unit 330 accepts an operation input performed by a user. The input is transferred to the control unit 310 as a command signal, and the control unit 310 can execute a predetermined control or calculation as necessary. The output unit 340 is a display or the like, and outputs the results of processing by the control unit 310, information input via the input unit 330, information received from the server device 100 via the communication unit 350, and the like. The communication unit 350 is a NIC or the like, which connects the client device 110 to the network 150 and controls communication with other devices.

[0024] 3. Functional configuration of the server device 100 The control unit 210 is configured to execute a receiving step, an acquiring step, a display control step, a generating step, and an analyzing step.

[0025] The reception step is a step of receiving a user's input to a search screen. For example, the first reception step is a step of receiving first information input by the user regarding the matter the user wishes to search. The second reception step is a step of receiving a selection of one particular theme among the themes of the literature. Details will be described later.

[0026] The acquisition step is a step of acquiring a plurality of keywords that are recalled from the matter through a large-scale language model using a predetermined prompt including the input first information, as will be described in detail later.

[0027] The display control step is a step of controlling visual information such as images including screens, still images or videos, icons, messages, etc. to be displayed on the output unit 340 of the client device 110. The display control step may generate only rendering information for displaying the visual information on the output unit 340 of the client device 110. Specifically, the first display control step is a step of displaying second information in which matters are organized by themes of documents based on a plurality of keywords and a database storing information on a plurality of documents. The second display control step is a step of displaying third information related to a selected specific theme. Details will be described later.

[0028] The first generation step is a step of generating a prompt to be input to the large-scale language model, and the second generation step is a step of generating second information and third information to be displayed on the client device 110. Details will be described later.

[0029] The analysis step is a step of performing analysis based on a plurality of keywords obtained from the large-scale language model and a database storing information on a plurality of documents, as will be described in detail later.

[0030] 4. Information processing flow In this section, an information processing method of the information processing system 1000 according to the first embodiment will be described. Fig. 4 is a sequence diagram (part 1) showing an example of information processing in the information processing system 1000. Fig. 5 is a diagram showing an example of a screen 500. Fig. 6 is a sequence diagram (part 2) showing an example of information processing in the information processing system 1000. Fig. 7 is a diagram showing an example of a screen 700. Fig. 8 is a diagram showing an example of a screen 800.

[0031] As described below, the information processing system 1000 includes at least one processor. The processor includes the following steps. The program of this embodiment is a program for executing each step of the information processing method of the information processing system 1000. The order of the processes can be changed as appropriate, multiple processes can be executed simultaneously, or some processes can be omitted.

[0032] 4.1 Flow Overview (Part 1) 4 is a sequence diagram (part 1) showing an example of information processing in the information processing system 1000. A user inputs a query related to an item that the user wishes to search for via a search screen displayed on the client device 110, and the server device 100 executes processing. Each sequence in this sequence diagram will be described below.

[0033] First, in sequence SQ401, control unit 310 of client device 110 transmits a display request for a search screen to server device 100 based on a user operation or the like.

[0034] Next, in sequence SQ402, the control unit 210 of the server apparatus 100 transmits a search screen to the control unit 310 of the client apparatus 110 based on a request from the control unit 310 of the client apparatus 110.

[0035] Next, in sequence SQ403, control unit 310 displays the received search screen on output unit 340. The search screen has an area where the user can input items that he or she wishes to search for.

[0036] 5 is a diagram showing an example of a screen 500. The screen 500 includes an input field 510 and an output field 520. The search screen may be the screen 500. In sequence SQ403, the screen 500 is displayed with the input field 510 and the output field 520 blank. The input field 510 is an area in which a user inputs a query relating to a matter he or she wishes to search for. The output field 520 is an area in which search results corresponding to the query input in the input field 510 are output. The output field 520 will be described later.

[0037] The user inputs a query (an example of first information) related to a matter to be searched for in the input field 510. For example, if the user wants to know information related to "healthcare by intestinal bacteria", the user inputs "healthcare by intestinal bacteria" in the input field 510 as a query.

[0038] Next, in sequence SQ404, the control unit 210 of the server device 100 accepts a query input in the input field 510 via the client device 110. Here, the control unit 210 can accept any text via the input field 510. That is, the query may be a two-word sentence consisting of two words, such as "healthcare by intestinal bacteria", or may be a single word or a sentence. Such a word or phrase input as a query is also called a search word. In other words, the query, which is an example of the first information related to the matter that the user wants to search for, includes information indicating a search word. However, it is preferable that the query includes information indicating a technical field, such as "healthcare by intestinal bacteria", rather than only "intestinal bacteria". By including information indicating a technical field in the query, the matter that the user wants to search for becomes clearer than when the query is only a word, and the information that the user wants to know can be obtained more specifically. In other words, the query, which is an example of the first information, may include information indicating a technical field.

[0039] Next, in sequence SQ405, the control unit 210 generates a prompt to be input to the large-scale language model. The prompt includes a query input by the user. The prompt is a request statement indicating an instruction to the large-scale language model and rules and conditions for executing the instruction.

[0040] Specifically, the prompt includes an instruction to output multiple keywords that are recalled from the input query. As a generating step, the control unit 210 generates a prompt according to the input query. Furthermore, the control unit 210 may generate a prompt that includes a purpose for outputting multiple keywords and rules for extracting multiple keywords. By inputting a prompt that clearly indicates the purpose and rules to the large-scale language model, the control unit 210 can acquire multiple keywords that are more suitable for the purpose from the large-scale language model. That is, as a generating step, the control unit 210 generates a prompt that instructs outputting multiple keywords for finding literature related to the query.

[0041] Here, the control unit 210 may generate a prompt to present a plurality of keywords from general knowledge related to the query as the generating step. In this manner, it is possible to generate information that is easy to understand for users other than researchers who specialize in the field.

[0042] Furthermore, the control unit 210 may generate, as a generating step, a prompt that instructs the user to classify the relationship between the multiple keywords and the query, in particular the relationship between the multiple keywords and the query. Also, the control unit 210 may generate, as a generating step, a prompt that instructs the user to output summary information of the multiple keywords related to the query.

[0043] Next, in sequence SQ406, the control unit 210 transmits the generated prompt to the large-scale language model of the information processing system 120.

[0044] Next, in sequence SQ407, the information processing system 120 receives a prompt from the control unit 210 and inputs the prompt as input data to the large-scale language model. The large-scale language model outputs a plurality of keywords based on the prompt.

[0045] Next, in sequence SQ408, the control unit 210 acquires a plurality of keywords output from the large-scale language model. In other words, as an acquisition step, the control unit 210 acquires a plurality of keywords that are reminiscent of the matter that the user wants to search for, which is input as a query, via the large-scale language model, using a predetermined prompt including the input query (an example of first information).

[0046] Next, in sequence SQ409, the control unit 210 generates output information that organizes the items that the user wants to search for by document theme based on a plurality of keywords and a document database that stores information on a plurality of documents. The output information is organized by document theme based on information on documents related to keywords. For example, the control unit 210 inputs a plurality of keywords into a database and analyzes them to generate output information that organizes the items that the user wants to search for, which are input as a query.

[0047] To explain the process of sequence SQ409 more specifically, the control unit 210 inputs a plurality of keywords received from the large-scale language model into a literature database and outputs information on literature related to each of the plurality of keywords. The control unit 210 can generate output information in which the matters that the user wants to search for are organized for each theme of the literature by using a trained model capable of performing classification and visualization in the literature database. As the trained model, an existing trained model such as a Word Embedding model, BERT, or LDA (Latent Dirichlet Allocation) can also be used. For example, the control unit 210 can generate a summary of the literature information by using the above-mentioned literature database or trained model, analyze the generated summary to identify the theme of the literature information, and classify the literature information by theme. Furthermore, the control unit 210 can generate output information in which the matters that the user wants to search for are organized based on the literature information classified by theme.

[0048] Another example of a method for obtaining output information will be described. When generating a prompt in sequence SQ405, the control unit 210 may generate a prompt that instructs the control unit 210 to classify the relationship between a plurality of keywords and the query, particularly the semantic relationship between the plurality of keywords and the query. By generating such a prompt, the control unit 210 can obtain information on the relationship between the query and the plurality of keywords from a large-scale language model in sequence SQ408. This allows the control unit 210 to generate output information that organizes the items that the user wants to search by matching the relationship between the query and the plurality of keywords with literature information obtained based on the plurality of keywords.

[0049] Furthermore, in sequence SQ409, control unit 210 may execute a combination of the above methods. By executing the process of sequence SQ409, control unit 210 can generate output information that is easy to understand for users other than researchers in specialized fields. That is, control unit 210 generates output information (second information) to be displayed on client device 110 as a generation step.

[0050] Next, in sequence SQ410, the control unit 210 transmits the generated output information to the client device 110 and controls the client device 110 to display it. As a first display control step, the control unit 210 displays output information (second information) that organizes items that the user wants to search for, which are input as a query for each theme of multiple documents, based on multiple keywords and a database that stores information on multiple documents.

[0051] Next, in sequence SQ411, the control unit 310 displays the output information on the screen. An example of this output information is displayed in the output field 520 of FIG. 5. In the example of FIG. 5, an overhead view is displayed when "robotics" is input as a query (which is a single word and can also be called a search word). The overhead view is an example of a diagram that visualizes the relationship between the themes of the literature. That is, the output information (second information) may be a diagram that visualizes the relationship between the themes of the literature. The circles indicate each of a plurality of themes related to technology. The circles are an example of objects that indicate a plurality of themes related to technology. The shape of the object is not limited to a circle and may be other shapes. The size of the circle indicates the amount of literature related to the keyword. In addition, the lines connecting the circles indicate the relevance of the technical fields. In addition, the closer the distance between the circles is, the higher the relevance of the technical fields is, and the farther the distance is, the lower the relevance of the technical fields is. Note that the overview in Figure 5 is an overview when "robotics" is entered as a query, so the diagram visualizing the relationships between the themes of the literature can also be said to be a diagram visualizing the relationships between the technical fields. Furthermore, literature related to robotics can also be said to be technical documents.

[0052] In this manner, a user can obtain information that visualizes multiple themes and the relevance of each theme related to literature containing keywords related to the query, simply by inputting a technology-related query. In the bird's-eye view, the amount of literature is represented by the size of the circle, the relevance is represented by the lines connecting the circles, and the relevance of the technical fields is represented by the distance between the circles. Therefore, a user can first refer to the bird's-eye view to get an idea of ​​which fields of literature he or she should read.

[0053] The output information may be a list of literature by topic related to a query regarding a matter that the user wants to search for. In this manner, the user can check a list of literature by topic.

[0054] To sum up, the control unit 210 of the server device 100 receives a query (first information) input by a user regarding an item the user wishes to search for as a first receiving step. The control unit 210 acquires a plurality of keywords associated with the item through a large-scale language model using a predetermined prompt including the input first information as an acquisition step. The control unit 210 displays output information (second information) in which items are organized by theme of a plurality of documents based on a plurality of keywords and a database storing information on a plurality of documents as a first display control step.

[0055] According to this aspect, when a researcher (an example of a user) wants to understand information in a field other than his / her specialty, the information processing system 1000 can obtain an overview map or a literature list by theme (an example of second information) that is easy to understand even for a researcher outside his / her specialty when the researcher inputs information on the subject to be searched through the client device 110. That is, when a researcher wants to understand information in a field other than his / her specialty, he / she can get an idea of ​​important themes and literature related to the subject to be searched without having to read all the literature with advanced content. As a result, the time required for hypothesis construction and preliminary research in business development using science and technology can be shortened. In addition, by using a large-scale language model such as GPT, keywords that are easy for people to see and understand can be generated instead of difficult keywords, and the keywords can be used to display an overview map, a literature list by theme, etc.

[0056] 4.2 Flow Overview (Part 2) Fig. 6 is a sequence diagram (part 2) showing an example of information processing in the information processing system 1000. The information processing shown in Fig. 6 is information processing performed after the information processing shown in Fig. 4. The following will be described along with each sequence of this sequence diagram.

[0057] First, in sequence SQ601, control unit 210 determines whether or not a predetermined operation has been performed on screen 500 in Fig. 5. An example of the predetermined operation is an operation of selecting a circle included in the bird's-eye view of Fig. 5. When a circle is selected on a screen such as that shown in Fig. 5, operation information including the fact that the circle has been selected and information about the selected circle is transmitted from client device 110 to server device 100. When control unit 210 receives such operation information, it determines that a predetermined operation has been performed on screen 500 in Fig. 5. Control unit 210 may accept a selection of one specific theme from among the themes of the literature as a second acceptance step.

[0058] If it is determined that a predetermined operation has been performed, then in sequence SQ602, the control unit 210 generates third information to be displayed on the client device 110 as a generation step. For example, the control unit 210 generates a screen 700 including structured information as shown in FIG. 7, which will be described later. Here, the control unit 210 generates the screen 700 including the structured information by referring to a literature database stored in the storage unit 220. The control unit 210 stores information related to a plurality of keywords acquired from the large-scale language model and information organized by literature theme using the literature database based on the information in the literature database together with a plurality of pieces of literature information, and causes the storage unit 220 to store the information.

[0059] Next, in sequence SQ603, the control unit 210 transmits the generated screen to the client device 110. That is, the control unit 210 may display structured information, which is an example of third information related to the selected specific theme, as a second display control step.

[0060] Next, in sequence SQ604, control unit 310 displays the generated screen on output unit 340.

[0061] FIG. 7 is a diagram showing an example of a screen 700. The screen 700 includes an output field 710, an output field 730, an output field 760, and a selection field 770. The output field 710 includes structured information. The structured information is information that structures a field selected by a user from fields related to the literature displayed on the screen 500 (an example of the theme of the literature). The example of FIG. 7 shows structured information when the field of Computer science is selected on the screen of FIG. 5. More specifically, it shows that themes such as Image Recognition, SVM, Classification, and Optimization exist in a lower hierarchy of the field of Classification.

[0062] In the example of FIG. 7, a selection object 720, which is an example of an object for selecting a user's expertise level, is displayed near the structured information. When the user operates the selection object 720 to select his / her own expertise level, the control unit 210 changes the structured information according to the user's expertise level selected by the user. In other words, the control unit 210 displays structured information (third information) according to the user's knowledge level as a second display control step. In the example of FIG. 7, the expertise levels are shown in three stages, 0, 1, and 2, but are not limited to these. It is assumed that the expertise level is 0, which is the lowest, and 2, which is the highest. For example, the control unit 210 generates structured information using words in a specialized field as the expertise level increases, and generates structured information using words that are easy to understand even for non-experts as the expertise level decreases. It is an example to change the understandability according to the expertise level. As another example, the control unit 210 may generate structured information with a finer hierarchical structure as the expertise level increases, and generate structured information with a coarser hierarchical structure as the expertise level decreases.

[0063] The structured information (third information) may include statistical information on literature related to a particular theme. Output field 730 includes statistical information on literature in a field selected by the user from among fields related to literature displayed on screen 500. Here, an example of literature is a research paper. If the literature is a research paper, graph 740 is the number of research papers published by year in the selected field in a specified period. Graph 750 is the number of research papers published by country in a specified year in the selected field. The number of research papers published by year in a specified period in the selected field and the number of research papers published by country in a specified year in the selected field are examples of statistical information.

[0064] The output field 760 includes information on technical terms in a field related to the field selected by the user among the fields related to the documents displayed on the screen 500. For example, the technical terms in the field related to the selected field are output to the output field 760 in the fields of Image Recognition, SVM, Classification, and Optimization, which are related to the selected field Computer science, with reference to FIG. 7 as an example. The control unit 210 may display, in the output field 760, summaries of important papers with the highest number of citations among the fields related to the selected field. The control unit 210 may display a selection object that allows the user to select what to display in the output field 760. The control unit 210 may output information selected by the selection object to the output field 760. As described above, information output to the output field 760 includes explanations of technical terms and summaries of papers. Although only one output field 760 is indicated by a reference number in the example of FIG. 7, four fields having the same shape are the same output field. The number of output fields that display explanations of technical terms and summaries of papers is not limited to four.

[0065] The selection field 770 is a field for selecting the display language in Fig. 7. When Japanese is selected in the selection field 770, explanations of technical terms, etc. are displayed in Japanese, and when English is selected in the selection field 770, explanations of technical terms, etc. are displayed in English.

[0066] That is, the control unit 210 controls to display an overhead view in which a plurality of technical fields are clustered and the relevance of each technical field is visualized. The control unit 210 controls to display structured information, which is information that structures a field selected by a user from among fields related to documents displayed on the screen, based on a user operation via a screen including the overhead view.

[0067] Users can simply enter a search term related to the technology to obtain information that visualizes multiple themes related to documents that contain keywords related to the search term, as well as the relationships between each theme. In addition, by selecting a theme in the overview view, structured information on the selected theme can be obtained.

[0068] In addition, based on user operation via screen 500 including the overhead view, control unit 210 controls to display screen 700 including structured information, which is information that structures a theme selected by the user from the themes related to the literature displayed on screen 500, statistical information on the literature on the theme selected by the user, and information on technical terms related to the theme selected by the user. By selecting a theme from the bird's-eye view, the user can visually view, all at once on the same screen, the structured information for the selected theme, statistical information for the literature on the theme, and information on technical terms related to the theme.

[0069] Next, in sequence SQ606, control unit 210 determines whether or not a predetermined operation has been performed on screen 700 in Fig. 7 or screen 500 in Fig. 5. Examples of the predetermined operation include an operation in which a theme or a technical term corresponding to the theme is selected in the structured information in Fig. 7, or an operation in which an object included in the overhead view in screen 500 in Fig. 5 is double-tapped. When control unit 210 receives information indicating that such an operation has been performed from client device 110, control unit 210 determines that a predetermined operation has been performed on screen 700 in Fig. 7 or screen 500 in Fig. 5.

[0070] If it is determined that a predetermined operation has been performed, in sequence SQ607, control unit 210 generates a screen 800 as shown in FIG. 8, which will be described later.

[0071] Next, in sequence SQ608, the control unit 210 transmits the generated screen to the client device 110.

[0072] Next, in sequence SQ609, the control unit 310 displays the screen 800 on the output unit 340.

[0073] 8 is a diagram showing an example of a screen 800. The screen 800 includes information related to a structured theme indicated by structured information selected by a user operation, or a technology corresponding to the theme. Alternatively, the screen 800 includes information related to a theme corresponding to an object included in the overhead view selected by the user. That is, the control unit 210 controls to display related information related to the technology of the structured theme indicated by the structured information based on a predetermined user operation. The related information of the structured information is also an example of the third information. The control unit 210 switches the display of related information by the selection of a tab by the user. For example, when the tab 810 is selected, the control unit 210 controls to display, as related information, information on a startup company that has been established using the technology of the selected theme on the screen 800. The information on the startup company includes, for example, the company name of the startup company, access information to the company's homepage, the total amount obtained through fundraising, the number of employees, the year of establishment, information on patents held, and overview information of the company. Also, for example, when the tab 820 is selected, the control unit 210 controls to display, as related information, information on patents that have been published within a predetermined period related to the technology of the selected theme on the screen 800. The predetermined period is, for example, a period of the most recent three years, and may be determined in advance or may be set by the user via a screen or the like. The information about the patent includes, for example, the patent number, access information of the patent publication, information about the lead inventor, information about the right holder, information about the expiration date of the patent term, and information about the outline of the patent.

[0074] (Variation 1) A first modified example of the embodiment will be described.

[0075] In sequence SQ404, the control unit 210 may accept input of a search word and selection of a technical field via the input field 510 of the screen 500. The control unit 210 causes a plurality of technical fields to be displayed on the screen 500 in a selectable manner. Examples of technical fields include physics, computer science, mathematics, chemistry, and artificial intelligence. According to this embodiment, in sequence SQ405, the control unit 210 can generate a prompt according to the selected technical field.

[0076] In sequence SQ406, the control unit 210 may input a prompt to the large-scale language model of the information processing system 120 via an API (Application Programming Interface).

[0077] Furthermore, it may be provided in the following aspects:

[0078] (1) An information processing system comprising at least one processor configured to execute each of the following steps: a first reception step receives first information input by a user relating to a matter the user wishes to search for; an acquisition step acquires a plurality of keywords evoked by the matter via a large-scale language model using a predetermined prompt including the input first information; and a first display control step displays second information in which the matters are organized by theme of the plurality of documents based on the plurality of keywords and a database storing information on a plurality of documents.

[0079] According to this embodiment, it is possible to display output information in which the items input by the user are organized for each document theme.

[0080] (2) In the information processing system described in (1) above, the second information is a diagram that visualizes the relationships between the themes of the multiple documents.

[0081] According to this aspect, it is possible to organize the input items by document theme and display output information that visualizes the relationships between the documents' themes.

[0082] (3) In the information processing system described in (1) or (2) above, the second information is a list of literature for each theme related to the matter.

[0083] According to this aspect, the input items can be organized by themes of the documents, and a list of documents organized by theme can be displayed as output information.

[0084] (4) In the information processing system according to any one of (1) to (3) above, the first information includes information indicating a technical field.

[0085] According to this aspect, it is possible to display output information that organizes items within a technical field by using a query that includes information indicating the technical field.

[0086] (5) In the information processing system according to any one of (1) to (4) above, the first information includes information indicating a search word.

[0087] According to this aspect, when the user inputs information including a word to be searched for, output information can be displayed that is organized according to the theme of the document with respect to information including the input word.

[0088] (6) In the information processing system described in any one of (1) to (5) above, a second reception step further receives a selection of one specific theme from the themes of the plurality of documents, and a second display control step further displays third information related to the selected specific theme.

[0089] According to this aspect, it is possible to display information related to a specific theme from among the output information in which the items input by the user are organized for each document theme.

[0090] (7) In the information processing system described above in (6), in the second display control step, the third information is displayed according to a knowledge level of the user.

[0091] According to this embodiment, it is possible to display information according to the knowledge level of the user.

[0092] (8) In the information processing system described in (6) or (7) above, the third information includes at least one of statistical information of the plurality of documents related to the specific theme, explanations of technical terms, summaries of the plurality of documents, information on startup companies, and patent information.

[0093] According to this embodiment, it is possible to display statistical information on literature, explanations of technical terms, summaries of literature, information on start-up companies, patent information, and the like, as information related to a particular theme.

[0094] (9) An information processing program that causes a computer to execute each step of the information processing system described in any one of (1) to (8) above.

[0095] According to this embodiment, it is possible to display output information in which the items input by the user are organized for each document theme.

[0096] (10) An information processing method, comprising steps executed by an information processing system according to any one of (1) to (8) above.

[0097] According to this embodiment, it is possible to display output information in which the items input by the user are organized for each document theme. Of course, this is not limiting. Furthermore, the following aspects are also possible.

[0098] <1> An information processing system comprising: a search word input via a screen and field information indicating a plurality of technical fields; a request to a large-scale language model to generate keywords relating to technical documents related to the search word in each of the plurality of fields; the keywords generated in the large-scale language model; and control to display bird's-eye view information that is clustered into a plurality of technical fields based on information on technical documents related to the keywords and in which the relevance of each technical field is visualized.

[0099] <2> the above <1> When a technical document including the keyword exists, the information processing system controls to display the overhead view information using information of the technical document including the keyword.

[0100] <3> the above <1> or <2> In the information processing system described in the above, it is determined whether or not a technical document containing the keyword exists, and if a technical document containing the keyword exists, the information of the technical document containing the keyword is used to control the display of the overhead view information.

[0101] <4> the above <1> from <3> In the information processing system described in any one of the above, based on a user operation via a screen including the overhead view information, the information processing system controls to display structured information which is information that structures a field selected by a user from among fields related to the technical document displayed on the screen.

[0102] <5> the above <4> In the information processing system according to the above, the structured information is changed according to a level of expertise of the user.

[0103] <6> the above <1> from <5> In the information processing system described in any one of the above, based on a user operation via a screen including the overhead view information, the information processing system controls to display statistical information of technical documents in a field selected by a user from among fields related to the technical documents displayed on the screen.

[0104] <7> the above <1> from <6> In the information processing system described in any one of the above, based on a user operation via a screen including the overhead view information, the information processing system controls to display information regarding technical terms in a field related to a field selected by a user from among fields related to the technical document displayed on the screen.

[0105] <8> the above <1> from <7> and based on a user operation via a screen including the bird's-eye view information, controls to display a screen including structured information, which is information that structures a field selected by a user from among fields related to technical documents displayed on the screen, statistical information on technical documents in the field selected by the user, and information on technical terms related to the field selected by the user.

[0106] <9> the above <4> or <5> or <8> In the information processing system described in the above, the information processing system controls to display related information related to the technology in the structured field indicated in the structured information based on a predetermined user operation.

[0107] <10> the above <9> In the information processing system described in the above, the related information is information about a startup that was established using technology in the field.

[0108] <11> the above <9> In the information processing system described in the above, the related information is information regarding patents published within a specified period of time relating to technology in the field.

[0109] <12> the above <1> from <11> In any one of the information processing systems described above, the technical document is a research paper on a technology or a patent document on a technology.

[0110] <13> An information processing system that controls the display of bird's-eye view information that is clustered into a plurality of technical fields and in which the relevance of each technical field is visualized, and controls the display of structured information that is information that structures a field selected by a user from among the fields related to technical documents displayed on the screen based on user operation via a screen including the bird's-eye view information.

[0111] <14> An information processing method executed by an information processing system, comprising: making a request to a large-scale language model to generate keywords related to technical documents related to the search word in each of a plurality of technical fields, the keywords being included in a search word input via a screen and field information indicating a plurality of technical fields; acquiring the keywords generated in the large-scale language model; and controlling the display of bird's-eye view information that is clustered into a plurality of technical fields based on information on technical documents related to the keywords and in which the relevance of each technical field is visualized.

[0112] <15> An information processing method executed by an information processing system, the information processing method controlling the display of bird's-eye view information that is clustered into a plurality of technical fields and in which the relevance of each technical field is visualized, and controlling the display of structured information that is information that structures a field selected by a user from among the fields related to technical documents displayed on the screen based on user operation via a screen including the bird's-eye view information.

[0113] <16> A program for executing the above-mentioned <1> from <13> A program for executing the information processing system according to any one of claims 1 to 5.

[0114] For example, a part or all of the processes of the server device 100 described above may be executed by the client device 110 .

[0115] Finally, although various embodiments of the present invention have been described, these are presented as examples and are not intended to limit the scope of the invention. The new embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. The embodiments and their modifications are within the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0116] 100: Server device 110: Client device 120: Information Processing Systems 150: Network 210: Control section 220: Storage section 230: Communications Department 310: Control section 320: Storage section 330: Input section 340: Output section 350: Communications Department 500: Screen 510: Input field 520 : Output field 700: Screen 710 : Output field 720 : Selected object 730 : Output field 740: Graph 750: Graph 760: Output field 770: Selection field 800: Screen 810: Tab 820: Tab 1000: Information processing systems

Claims

1. An information processing system, At least one processor; The processor is configured to perform the steps of: In the first receiving step, first information input by a user regarding a matter that the user wishes to search is received; In the acquisition step, a plurality of keywords associated with the item are acquired through a large-scale language model using a predetermined prompt including the input first information; In the first display control step, the system displays second information that organizes the items by theme of the multiple documents based on the multiple keywords and a database that stores information on the multiple documents.

2. 2. The information processing system according to claim 1, The system, wherein the second information is a diagram that visualizes the relationships between themes of the multiple documents.

3. 2. The information processing system according to claim 1, The system, wherein the second information is a list of literature for each of the topics related to the subject.

4. 2. The information processing system according to claim 1, The first information includes information indicating a technology field.

5. 2. The information processing system according to claim 1, The system, wherein the first information includes information indicative of a search word.

6. 2. The information processing system according to claim 1, Furthermore, in a second receiving step, a selection of one specific theme among the plurality of themes of documents is received; Furthermore, in the second display control step, the system displays third information related to the selected specific theme.

7. 7. The information processing system according to claim 6, In the second display control step, the third information is displayed according to a knowledge level of the user.

8. 7. The information processing system according to claim 6, The third information is related to the specific theme. statistical information of the plurality of documents; Explanation of technical terms and A summary of said documents; Information on startup companies, Patent information and A system comprising at least one of the following:

9. An information processing program, A program for causing a computer to execute each step of the information processing system according to any one of claims 1 to 8.

10. 1. An information processing method, comprising: A method comprising the steps executed by an information processing system according to any one of claims 1 to 8.

Citation Information

Patent Citations

  • Theme prompt type keyword extraction method based on text semantic enhancement

    CN114970523A

  • Recommending topic clusters for unstructured text documents

    US20180032606A1

  • Information processing apparatus

    JP2010262519A

Cited By

  • Response output device

    WO2026204003A1