Information processing device, program, and knowledge graph generation method

Abstract

To provide an information processing device that generates a knowledge graph that is easy for people to see and understand and that leads to the discovery of new knowledge. [Solution] The information processing device 1 includes a first control unit 151 that outputs a first knowledge graph generated by analyzing a sentence, a second control unit 152 that outputs information indicating episodes associated with episode nodes included in the first knowledge graph, and a third control unit 153 that outputs a second knowledge graph generated based on the first knowledge graph, the second knowledge graph including at least two nodes selected from place nodes that are nodes related to locations, person nodes that are nodes related to people, object nodes that are nodes related to objects other than people, and event nodes that are nodes related to events, and an edge connecting the two nodes. The at least two nodes are generated from episode nodes included in the first knowledge graph, and the edge connecting the two nodes is generated from a time-series edge.

Description

[Technical Field]

[0001] The present disclosure relates to techniques for generating knowledge graphs. [Background technology]

[0002] A knowledge graph is a network of knowledge that systematically connects various pieces of knowledge and is represented by a graph structure consisting of nodes and edges. A knowledge graph is also called a knowledge graph.

[0003] For example, Patent Document 1 discloses a knowledge graph support system that supports the understanding of knowledge graphs based on natural language generation technology in order to solve the problem that conventional knowledge graphs generate knowledge graphs that are difficult to understand due to a lack of quality control. According to Patent Document 1, it is possible to support experts in the relevant field in understanding knowledge graphs and performing quality control of knowledge graphs. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Permit No. 7064262 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the knowledge graph support system disclosed in Patent Document 1 merely supports understanding of the knowledge graph by converting the generated knowledge graph into natural language text using natural language generation technology, and does not solve the problem of generating knowledge graphs that are difficult to understand.

[0006] In addition, although the generated knowledge graph can generally show how information (knowledge) is related and combined with other information (knowledge), as the amount of information (knowledge) increases, the knowledge graph becomes more complex, making it difficult for people to read through the information (knowledge) and use it.

[0007] In one aspect, the present disclosure aims to provide an information processing device and the like that can generate a knowledge graph that is easy for people to see and understand and that leads to the discovery of new knowledge. [Means for solving the problem]

[0008] (1) An information processing device of the present disclosure is an information processing device that generates a knowledge graph composed of a plurality of nodes and edges indicating the relationships between each of the plurality of nodes. This information processing device includes: a storage unit that stores a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating history included in the sentence; a first control unit that controls output of the first knowledge graph from the storage unit in a manner in which the episode nodes included in the first knowledge graph are connected by the history edges; a reception unit that receives instructions from a user; and a third unit that outputs information indicating episodes associated with the episode nodes included in the output first knowledge graph based on the instructions received by the reception unit. and a third control unit that controls the output of a second knowledge graph generated based on the first knowledge graph, the second knowledge graph including at least two nodes selected from a place node which is a node related to a location, a person node which is a node related to a person, an object node which is a node related to an object other than a person, and an event node which is a node related to an event, and an edge connecting the two nodes, the second knowledge graph being generated based on the first knowledge graph in a manner in which the two nodes are connected by the edge, wherein the at least two nodes are generated from the episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from the history edge.

[0009] The "first knowledge graph" is a knowledge graph (episode graph) generated in which each of the multiple episodes contained in a sentence is a node (episode node), and each of the episode histories contained in the sentence is an edge (history edge).

[0010] The "second knowledge graph" is a knowledge graph (human-thing-word graph) consisting of place nodes, which are nodes related to locations, person nodes, which are nodes related to people, thing nodes, which are nodes related to things other than people, and event nodes, which are nodes related to events, and edges connecting the two nodes. The second knowledge graph is generated based on the first knowledge graph.

[0011] "Text" refers to text that includes multiple episodes, such as history. Note that the text here is not limited to historical texts, but may also be, for example, a recipe or a statement included in a meeting minutes.

[0012] (2) In such an information processing device, the process edges may include at least one of a physical process edge indicating a process based on the passage of time or a procedure, and a logical process edge indicating a process based on a causal relationship.

[0013] "Background" is a concept that indicates the physical background or logical background of the four types of relationships between episodes (identical or similar relationships, physical background relationships, logical background relationships, and relationships that indicate complete dissimilarity).

[0014] (3) The system may further include a generation unit that performs a 5W1H analysis of the text, defines sentences containing the elements of 4W (When, Where, Who, What) as episodes, and defines sentences containing the elements of 1W1H (Why, How) as contexts, and generates the episode nodes and the context edges.

[0015] · "5W1H analysis" means organizing a text with the 5W1H elements of "When", "Where", "Who", "What", "Why", and "How". Through 5W1H analysis, a sentence composed of at least one or more of the 5W1H elements can be obtained.

[0016] (4) Also, the text shows a story related to history or legend. The person node is a node representing the entity that is the subject of the episode. The place node is a node representing the location related to the episode. The object node is a node representing the object related to the episode. The event node may be a node representing the event related to the episode. · "Place" represents a location, such as a historical or legendary spot, area, etc. · "Object" represents an object related to the episode, such as a cultural relic or a work in history or legend. · "Person" represents the subject of the episode, such as a person, a monster, an animal, a legendary existence, etc. · "Event" represents an event related to the episode, such as a historical or legendary event, incident, etc.

[0017] (5) In such an information processing device, when the reception unit receives a designation of a predetermined node among the nodes included in the second knowledge graph, it may further include a fourth control unit that outputs an explanation of the episode associated with the predetermined node and an explanation of the relationship between the predetermined node and other nodes connected by the edge.

[0018] (6) Also, the episode includes time information when the event occurred. When a plurality of episodes are associated with the predetermined node, the fourth control unit may further display the plurality of episodes in chronological order based on the time information.

[0019] (7) A program disclosed herein is an information processing device that generates a knowledge graph composed of a plurality of nodes and edges indicating associations between each of the plurality of nodes, the program including: a first control step of controlling output of a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating history included in the sentence, from a storage unit that stores the first knowledge graph, in a manner in which the episode nodes included in the first knowledge graph are connected by the history edges; a reception step of receiving an instruction from a user; and a first control step of controlling output of the episode nodes included in the output first knowledge graph based on the instruction received in the reception step. a second control step of outputting information indicating an episode associated with a node; and a third control step of controlling to output a second knowledge graph generated based on the first knowledge graph, the second knowledge graph including at least two nodes selected from a place node which is a node related to a location, a person node which is a node related to a person, an object node which is a node related to an object other than a person, and an event node which is a node related to an event, and an edge connecting the two nodes, the second knowledge graph being generated based on the first knowledge graph in a manner in which the two nodes are connected by the edge, wherein the at least two nodes are generated from the episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from the history edge.

[0020] (8) The method for generating a knowledge graph of the present disclosure is a method for generating a knowledge graph composed of a plurality of nodes and edges indicating the relationships between each of the plurality of nodes. This generation method includes a first control step of controlling output of a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating history included in the sentence, from a storage unit that stores the first knowledge graph, in a manner in which the episode nodes included in the first knowledge graph are connected by the history edges; a reception step of receiving an instruction from a user; and a second control step of outputting information indicating the episodes associated with the episode nodes included in the output first knowledge graph based on the instruction received in the reception step. and a third control step of controlling the output of a second knowledge graph generated based on the first knowledge graph, the second knowledge graph including at least two nodes selected from a place node which is a node related to a location, a person node which is a node related to a person, an object node which is a node related to an object other than a person, and an event node which is a node related to an event, and an edge connecting the two nodes, the second knowledge graph being generated based on the first knowledge graph in a manner in which the two nodes are connected by the edge, wherein the at least two nodes are generated from the episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from the history edge. [Effects of the Invention]

[0021] The information processing device and the like disclosed herein can generate knowledge graphs that are easy for people to see and understand and that lead to the discovery of new knowledge. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a functional block diagram of an information processing device according to the present disclosure. [Figure 2] FIG. 10 is a conceptual diagram illustrating an example of an episode graph. [Figure 3] FIG. 1 is a conceptual diagram showing an example of a human-monoword graph. [Figure 4]FIG. 2 is a functional block diagram of a user terminal of the present disclosure. [Figure 5] FIG. 1 is a schematic diagram illustrating an example of a hardware configuration of an information processing device. [Figure 6] FIG. 2 is a schematic diagram illustrating an example of a hardware configuration of a user terminal. [Figure 7] 10 is a schematic flow diagram illustrating an example of an operation of the information processing device. [Figure 8] 10 is a schematic flow showing an example of an operation for generating an episode graph. [Figure 9] FIG. 10 is a conceptual diagram for explaining a method for generating an episode graph. [Figure 10] FIG. 10 is a conceptual diagram for explaining a method for generating an episode graph. DETAILED DESCRIPTION OF THE INVENTION

[0023] An information processing device according to one embodiment of the present disclosure will be specifically described below with reference to the drawings. Note that each of the embodiments described below illustrates a specific example of the present disclosure. The numerical values, shapes, components, and component placement positions shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Furthermore, among the components in the following embodiments, components that are not described in an independent claim that represents a top-level concept will be described as optional components. Furthermore, the contents of each of the embodiments can be combined.

[0024] The drawings are schematic, and some components may be omitted for the sake of explanation. Components common to one or more embodiments may be designated by the same reference numerals, and explanations thereof may be omitted.

[0025] [1. Brief description]

[0026] Fig. 1 is a functional block diagram of an information processing device 1 of the present disclosure. Fig. 2 is a conceptual diagram showing an example of an episode graph. Fig. 3 is a conceptual diagram showing an example of a person-thing-word graph. Fig. 4 is a functional block diagram of a user terminal 2 of the present disclosure. First, an overview of the information processing device 1, the user terminal 2, etc. will be described.

[0027] (Information processing device) An information processing device 1 shown in FIG. 1 generates a knowledge graph made up of a plurality of nodes and edges indicating the relevance between each of the plurality of nodes.

[0028] The information processing device 1 generates a knowledge graph called an episode graph based on, for example, multiple historical episodes, and then generates a knowledge graph called a human-object-word graph based on the generated episode graph.The information processing device 1 then expresses multiple historical episodes using the generated human-object-word graph.The information processing device 1 can transmit a portion of the generated human-object-word graph to the user terminal 2.

[0029] (episode) An episode may be a story that specifically conveys a certain matter, or a fairly complete story, i.e., a short story. The episodes generated by the information processing device 1 are sentences expressed in the form of who took what action (what state occurred), and may be, for example, historical episodes, but are not limited to this. The episodes generated by the information processing device 1 may be, for example, one step in a cooking recipe, or the content of a statement made by one person in minutes of a meeting, etc.

[0030] (Episode graph) An episode graph is a knowledge graph made up of nodes (episode nodes) indicating each of the multiple episodes included in a sentence, and edges (chronological edges) indicating each of the episodes included in the sentence. An example of an episode graph is as shown in FIG. 2. That is, the episode graph is made up of episodes 1 to 7, episodes 1 and 2 are connected by a chronological edge 401, episodes 2 and 5 are connected by a chronological edge 402, and episodes 2 and 3 are connected by a chronological edge 403. Furthermore, episodes 2 and 4 are connected by a chronological edge 404, and episodes 3 and 5 are connected by a chronological edge 405. The chronological edges 401 to 405 are represented by directed line segments.

[0031] (Hitomonokotobagraph) A human-thing-word graph is a knowledge graph consisting of four types of nodes: place nodes, which are nodes related to locations; person nodes, which are nodes related to people; entity nodes, which are nodes related to entities other than people; and event nodes, which are nodes related to events; and edges connecting the nodes. It is generated based on an episode graph. An example of a human-thing-word graph is shown in FIG. 3. That is, the human-thing-word graph consists of person nodes 501, 502, and 503, place nodes 504 and 505, entity nodes 506 and 507, and entity node 508, with person node 501 and place node 504 connected by edge 601. Furthermore, place node 504 and entity node 506 are connected by edge 602, entity node 506 and place node 505 are connected by edge 603, place node 505 and person node 502 are connected by edge 604, and person node 502 and person node 503 are connected by edge 605. The person node 502 and the event node 508 are connected by an edge 606, the event node 508 and the entity node 507 are connected by an edge 607, and the event node 508 and the place node 504 are connected by an edge 608. Note that the edges 601 to 608 are represented by undirected line segments.

[0032] (User terminal 2) The user terminal 2 displays, for example, a knowledge graph (a part of the human-thing-word graph) transmitted from the information processing device 1. For example, when a node of the knowledge graph displayed on the user terminal 2 is specified (selected) by a user who is the user of the user terminal 2, the user terminal 2 displays an episode associated with the specified node. The user of the user terminal 2 can read the displayed episode and further specify a node connected by an edge to the node corresponding to the displayed episode, thereby proceeding to read the episode of the human-thing-word graph.

[0033] (Communication network) The information processing device 1 and the user terminal 2 are connected to each other so as to be able to communicate with each other via the Internet or an intranet (communication network). The communication network is, for example, the Internet. Any communication network using wired or wireless communication may be used. Examples of wireless communication include Bluetooth (registered trademark), Wi-Fi (registered trademark), and LPWA (Low Power Wide Area). LPWA may use LTE (registered trademark) or a frequency band that does not require a license.

[0034] [2.Detailed explanation] (Information processing device 1) The information processing device 1 is, for example, a computer such as a personal computer.

[0035] 1, the information processing device 1 includes an acquisition unit 11, an analysis unit 12, a generation unit 13, a storage unit 14, a control unit 15, and a reception unit 16. Note that the information processing device 1 does not necessarily have to include the analysis unit 12 or the storage unit 14. The storage unit 14 may be provided outside the information processing device 1 as long as it is usable by the information processing device 1.

[0036] (Information processing device 1, acquisition unit 11) The acquisition unit 11 acquires a text containing multiple episodes automatically or in response to an instruction from a user of the information processing device 1. For example, the acquisition unit 11 acquires a text such as a historical text via a communication network such as the Internet. Note that the text is not limited to historical texts, and may be any text containing multiple episodes, such as a cooking recipe or remarks included in minutes of a meeting.

[0037] (Information processing device 1, analysis unit 12) The analysis unit 12 analyzes the sentences acquired by the acquisition unit 11. More specifically, the analysis unit 12 performs a 5W1H analysis on the sentences acquired by the acquisition unit 11 and summarizes them into one or more sentences made up of 5W1H elements. Here, 5W1H analysis means organizing sentences using 5W1H elements such as "When," "Where," "Who," "What," "Why," and "How." The 5W1H analysis results in a sentence made up of at least one or more of the 5W1H elements.

[0038] The analysis unit 12 extracts, from the one or more aggregated sentences, sentences containing 4W elements as episodes, and sentences containing 1W1H elements as history. Here, 1W1H elements are elements such as "Why" and "How." Furthermore, from sentences containing 4W elements such as "When," "Where," "Who," and "What," it is possible to determine who took what action (the state that occurred), and to identify them as episodes. From sentences containing 1W1H elements, it is possible to determine what history occurred between episodes, and to connect the episodes. In this embodiment, the history indicated by sentences containing 1W1H elements is a history based on the passage of time or procedures (physical history), or a history based on causal relationships (logical history).

[0039] The relationships between episodes extracted by the analysis unit 12 are as follows: 1) Identical or similar relationships; 2) Relationships that show a time course or a procedural course, i.e., physical course relationships; 3) Relationships that show the course of events due to causal relationships, i.e., relationships of logical course; 4) Relationships that show that the two are completely different, i.e., relationships with different backgrounds. The analysis unit 12 can extract a physical process relationship or a logical process relationship from the four relationships between episodes as a process between episodes by extracting sentences containing 1W1H elements.

[0040] The analysis unit 12 may perform a 5W1H analysis of a sentence by issuing a prompt instruction using an AI such as a generation AI, and aggregate the sentences into one or more sentences composed of 5W1H elements. The analysis unit 12 may also perform a prompt instruction using an AI such as a generation AI, and extract, from the one or more aggregated sentences, sentences containing 4W elements as episodes and sentences containing 1W1H elements as history. In this case, such an AI may be provided outside the information processing device 1 and used by the analysis unit 12 via a communication network. Alternatively, such an AI may be provided in the information processing device 1 and used by the analysis unit 12.

[0041] (Information processing device 1, generation unit 13) The generation unit 13 generates an episode graph using episode nodes, which are nodes associated with the episodes extracted by the analysis unit 12, and history edges, which are edges associated with the history extracted by the analysis unit 12. The generation unit 13 stores the generated episode graph in the storage unit 14.

[0042] The generation unit 13 extracts, from the generated episode graph, episodes associated with episode nodes connected by time-series edges. The generation unit 13 extracts, from the extracted episodes, the entity that is the subject of the episode, a location related to the episode, an object related to the episode, or an event related to the episode. The generation unit 13 generates a person-monochrome-word graph using person nodes that represent the entity that is the subject of the episode, place nodes that represent locations related to the episode, object nodes that represent objects related to the episode, and event nodes that represent events related to the episode. The generation unit 13 stores the generated person-monochrome-word graph in the memory unit 14. In this way, the generation unit 13 generates nodes that constitute the person-monochrome-word graph from episode nodes included in the episode graph, and generates edges that constitute the person-monochrome-word graph from time-series edges included in the episode graph.

[0043] (Information processing device 1, storage unit 14) The storage unit 14 is configured with a hard disk drive (HDD), a solid state drive (SDD), a memory, etc. In this embodiment, for example, the storage unit 14 stores the analysis results of the analysis unit 12, and the episode graph and person-object-word graph generated by the generation unit 13.

[0044] (Information processing device 1, control unit 15) As shown in FIG. 1, the control unit 15 includes a first control unit 151, a second control unit 152, a third control unit 153, and a fourth control unit 154.

[0045] The first control unit 151 controls output of an episode graph from the storage unit 14 in a manner in which episode nodes included in the episode graph are connected by longitudinal edges. Specifically, the first control unit 151 outputs an episode graph such as that shown in FIG. 2, for example. The first control unit 151 may output and display the episode graph on a display connected to the information processing device 1. In the episode graph, as shown in FIG. 2, the episode nodes are displayed in a manner in which they each include an episode (episodes 1 to 7 in the figure) associated with the episode node, and the longitudinal edges are displayed as directed lines. In this way, the episode graph allows people to easily read through the episodes. Therefore, the episode graph can be a user interface that allows people to easily read through the episodes.

[0046] The second control unit 152 outputs information indicating an episode associated with an episode node included in the output episode graph, based on an instruction received by the reception unit 16. Specifically, the second control unit 152 outputs the content of the episode associated with the episode node received by the reception unit 16 and instructed by the user of the information processing device 1.

[0047] The third control unit 153 controls output of a human-monochrome-word graph generated based on the episode graph, the human-monochrome-word graph including at least two nodes selected from a place node, a person node, an object node, and an event node, and an edge connecting the two nodes. The third control unit 153 outputs the human-monochrome-word graph in a form in which at least two nodes are connected by an edge.

[0048] Specifically, the third control unit 153 outputs a human-object-word graph, for example, as shown in FIG. 3. The third control unit 153 may output and display the human-object-word graph on a display connected to the information processing device 1. In the human-object-word graph, as shown in FIG. 3, four types of nodes, namely, person nodes, thing nodes, event nodes, and place nodes, are displayed with icons that allow easy identification of their types, and edges are displayed with undirected lines. Furthermore, episodes associated with person nodes, thing nodes, event nodes, and place nodes are not displayed. In this way, the human-object-word graph classifies episodes into four types of nodes and can display the content of each episode, allowing people to easily read through the episodes. Therefore, the human-object-word graph can be a user interface that allows people to easily read through the episodes.

[0049] The fourth control unit 154 outputs an episode that corresponds to a specified node among the nodes included in the human-mono-word graph output by the third control unit 153, and that indicates the relationship between the specified node and other nodes connected by edges. In other words, by specifying any one of the person node, object node, event node, and place node that make up the human-mono-word graph, the description (content) of the episode that corresponds to the specified node and the content of the episode related to the specified node are displayed.

[0050] Note that, when an episode includes time information about when an event occurred and multiple episodes are associated with a specified node, the fourth control unit 154 may arrange and display the multiple episodes in chronological order or chronological order based on the time information. In other words, when multiple episodes are associated with a specified node and these episodes include time information, the multiple episodes are arranged and displayed in chronological order.

[0051] (Information processing device 1, reception unit 16) The reception unit 16 receives instructions from the user of the information processing device 1 or the user of the user terminal 2. In the present embodiment, the reception unit 16 can receive instructions from the user of the information processing device 1. Furthermore, the reception unit 16 can receive instructions from the user of the user terminal 2 transmitted from the user terminal 2. For example, the reception unit 16 may receive an instruction from the user of the information processing device 1 to specify an episode node from among the episode nodes included in the episode graph output by the first control unit 151. Furthermore, for example, the reception unit 16 may receive an instruction from the user of the information processing device 1 or the user terminal 2 to specify a predetermined node from among the nodes included in the human-thing-word graph output by the third control unit 153.

[0052] (User terminal 2) The user terminal 2 is a device, such as a smartphone, that can communicate with the information processing device 1. Note that the user terminal 2 may also be a tablet PC, a personal computer, or the like. In this embodiment, the user terminal 2 includes a display unit 21, a terminal acquisition unit 22, and an instruction transmission unit 23, as shown in FIG.

[0053] (User terminal 2, display unit 21) The display unit 21 displays information acquired by the user terminal 2 from the information processing device 1. As the display unit 21, for example, a device for displaying images such as a liquid crystal display (LCD), a plasma display panel (PDP), or an organic electroluminescence (EL) display is used. In this embodiment, the display unit 21 displays at least a part of the human-monoword graph acquired by the user terminal 2 from the information processing device 1. The display unit 21 can also display an explanation of an episode associated with a specific node designated by the user of the user terminal 2 among the nodes included in the human-monoword graph, and an episode indicating the relationship between the specific node and other nodes connected by edges.

[0054] (User terminal 2, terminal acquisition unit 22) The terminal acquisition unit 22 acquires information transmitted by the information processing device 1. In this embodiment, the terminal acquisition unit 22 acquires a person-thing-word graph and the like from the information processing device 1. The terminal acquisition unit 22 can acquire an episode that corresponds to a predetermined node designated by the user of the user terminal 2 from among the nodes included in the person-thing-word graph displayed on the display unit 21, and that indicates a description of the episode and the relationship between the predetermined node and other nodes connected by edges.

[0055] (User terminal 2, instruction sending unit 23) The instruction sending unit 23 sends an instruction from the user of the user terminal 2 to the information processing device 1. In this embodiment, the instruction sending unit 23 can send an instruction to the information processing device 1 to specify a specific node among the nodes included in the human-object word graph displayed on the display unit 21.

[0056] [3. Hardware configuration] 5 is a schematic diagram showing an example of the hardware configuration of the information processing device 1. FIG. 6 is a schematic diagram showing an example of the hardware configuration of the user terminal 2.

[0057] Next, the hardware configurations of the information processing device 1 and the user terminal 2 will be described with reference to FIGS.

[0058] (Hardware configuration of information processing device 1) The hardware configuration of the information processing device 1 will be described below. As shown in FIG. 5, the information processing device 1 of this embodiment uses, for example, a computer. The information processing device 1 includes a CPU 170. A memory 171, a connection port 173 for connecting / reading a storage device 172, etc., and a communication circuit 174 for communicating with the outside via a network are connected to the CPU 170 via a bus line 175. The memory 171 stores an information processing device program 1713 for processing the information processing device 1. A browser program 1712 and an OS 1711 (operating system) may also be stored. An episode graph management database (hereinafter referred to as episode graph DB) 1715 and / or a human-monochrome-word graph management database (hereinafter referred to as human-monochrome-word graph DB) 1716 may also be stored. The memory 171 may be the storage unit 14 or may be different from the storage unit 14. When the memory 171 and the storage unit 14 are different, the episode graph DB 1715 and the person-object-word graph DB 1716 are stored in the storage unit 14.

[0059] (Hardware configuration of user terminal 2) The hardware configuration of the user terminal 2 is almost the same as the hardware configuration of the above-described information processing device 1, so the same parts are denoted by the same reference numerals and the description thereof will be omitted. The user terminal 2 of this embodiment is, for example, a smartphone.

[0060] The computer's memory 171 stores a user terminal program 1717 for processing the user terminal 2. It may also store a browser program 1712 and an OS 1711 (operating system).

[0061] In this embodiment, the information processing device program 1713 and the user terminal program 1717 operate in cooperation with each other by utilizing the functions of the OS 1711 and the browser program 1712. Note that the information processing device program 1713 and the user terminal program 1717 may operate independently without utilizing the browser program 1712 and the OS 1711, respectively.

[0062] In the hardware configuration of the information processing device 1 and user terminal 2 described above, the functions shown in Figures 3 and 4 are realized, for example, using a CPU 170, a user terminal program 1717, and an information processing device program 1713, but some or all of them may be sequence-controlled using a logic circuit such as a microcomputer, or a PLC (programmable logic controller).

[0063] (Episode Graph DB1715) The episode graph DB 1715 stores episode graphs generated by the information processing device 1. The episode graph does not have to be in the form of a knowledge graph as shown in Fig. 2, for example, but may store elements that constitute the episode graph, such as a plurality of episode nodes, a plurality of history edges, and the connections between the plurality of history edges.

[0064] (Human and Mono Kotobuki Graph DB1716) The human-object-word graph DB 1716 stores a human-object-word graph generated by the information processing device 1. The human-object-word graph does not have to be in the form of a knowledge graph as shown in Fig. 3, for example, but it is sufficient if elements constituting the human-object-word graph, such as a plurality of nodes including place nodes, person nodes, thing nodes, and event nodes, a plurality of edges, and the connection relationships between the plurality of edges, are stored.

[0065] 4. Operation of Information Processing Device 1 (Operation of information processing device 1) FIG. 7 is a schematic flow diagram showing an example of the operation of the information processing device 1.

[0066] The operation of the information processing device 1 will be described below with reference to Fig. 7. Fig. 7 shows a flowchart illustrating one embodiment of the processing of the information processing device program 1713 used in the information processing device 1. The symbol "S" in the figure represents a step.

[0067] (S11) The CPU 170 of the information processing device 1 outputs the episode graph that was generated by analyzing the sentence and that is stored in the storage unit 14. In the present embodiment, the information processing device 1 outputs the episode graph that is stored in the storage unit 14 to a display connected to the information processing device 1, with the episode nodes connected by longitudinal edges. For example, an episode graph such as that shown in FIG. 2 is displayed on the display.

[0068] (S12) CPU 170 of information processing device 1 receives an instruction from the user. In the present embodiment, information processing device 1 receives an instruction from the user of information processing device 1 to specify an episode node from among the episode nodes included in the episode graph output in S11.

[0069] (S13) Based on the instruction received in S12, the CPU 170 of the information processing device 1 outputs information indicating the episode associated with the episode node. In this embodiment, the information processing device 1 outputs the content of the episode associated with the specified episode node. Note that the information processing device 1 may output and display an episode graph on a display connected to the information processing device 1.

[0070] (S14) The CPU 170 of the information processing device 1 outputs a person-thing-word graph generated based on the episode graph. In this embodiment, the information processing device 1 extracts episodes from the episode graph output in step S11, and extracts the entity that is the subject of the episode, a location related to the episode, an object related to the episode, or an event related to the episode as nodes. The information processing device 1 outputs a person-thing-word graph composed of a person node representing the entity that is the subject of the episode, a location node representing a location related to the episode, an object node representing an object related to the episode, and an event node representing an event related to the episode. Note that the information processing device 1 may output and display, for example, a person-thing-word graph such as that shown in FIG. 3 on a display connected to the information processing device 1.

[0071] Fig. 8 is a schematic flow showing an example of an episode graph generation operation of the information processing device 1. Figs. 9 and 10 are conceptual diagrams for explaining a method of generating an episode graph. Note that Fig. 10 is similar to Fig. 2.

[0072] 8 to 10, the operation of generating an episode graph by the information processing device 1 will be described below. Fig. 8 shows a flowchart illustrating an embodiment of the processing of the information processing device program 1713 used in the information processing device 1.

[0073] (S21) The CPU 170 of the information processing device 1 acquires a sentence including multiple episodes automatically or in response to an instruction from a user of the information processing device 1. For example, the information processing device 1 may acquire a sentence showing a story related to history or legend.

[0074] (S22) The CPU 170 of the information processing device 1 performs a 5W1H analysis of the text acquired in S21. For example, the information processing device 1 performs a 5W1H analysis of the text indicating a story related to history or legend acquired in S21, and aggregates the text into one or more sentences composed of 5W1H elements. The information processing device 1 may display the aggregated one or more sentences in bullet points on a display connected to the information processing device 1.

[0075] (S23) CPU 170 of information processing device 1 extracts sentences containing 4W elements as episodes based on the sentences analyzed using 5W1H in S22, and generates an episode node. For example, information processing device 1 may extract, as episodes, sentences containing 4W elements, such as what kind of people were active in what place, what cultural assets, works, or other objects were included, and what events, incidents, etc. occurred, from one or more sentences summarized using 5W1H analysis in S22. Then, information processing device 1 may generate an episode node associated with the extracted episode.

[0076] The information processing device 1 may display, for example, episode nodes 301 to 307 as shown in Fig. 9 on a display connected to the information processing device 1. The example shown in Fig. 9 shows a case where episodes 1 to 7 are extracted and episode nodes 301 to 307 are generated based on sentences that have been subjected to 5W1H analysis.

[0077] (S24) The CPU 170 of the information processing device 1 extracts sentences including 1W1H elements based on the sentences analyzed with 5W1H in S22, and generates a process edge. For example, the information processing device 1 may extract sentences including 1W1H elements that indicate what process (physical process or logical process) occurred between the episodes extracted in S23, and generate a process edge.

[0078] The information processing device 1 generates the longitudinal and longitudinal edges as directed segments, and multiple longitudinal and longitudinal edges connecting the same two episode nodes are generated together as a single longitudinal edge. Furthermore, from the viewpoint of making the episode graph easier to view, even if the information processing device 1 generates an autoregressive longitudinal edge (connecting to only one episode node), it does not display it on the display.

[0079] Furthermore, the information processing device 1 may display, for example, time and date edges 401 to 405 as shown in FIG. 10 on a display connected to the information processing device 1. In the example shown in FIG. 10, time and date edges 401 to 405, represented by directed lines, are generated based on sentences subjected to 5W1H analysis, and an episode graph is displayed in which the time and date edges 401 to 405 are added to episode nodes 301 to 307. In this manner, in the episode graph, episode nodes 301 to 307 associated with episodes 1 to 4 are related by time and date edges 401 to 405. Meanwhile, in the episode graph, episode nodes 306 and 307 associated with episodes 6 and 7 are nodes with unclear time and date, and exist as independent nodes with no time and date edges connected to them. Independent nodes are not used in human-thing-word graphs. Note that in the episode graph, multiple time and date edges are connected to a single episode node, such as episode node 305 associated with episode 5, which represents the existence of various different theories in history.

[0080] The information processing device 1 generates the person-thing-word graph shown in Fig. 3 based on the episode graph shown in Fig. 10. In this case, person nodes 501, 502, and 503 are nodes representing the subjects of the episode, such as people, monsters, animals, and legendary beings. Place nodes 504 and 505 are nodes representing locations, such as historical or legendary spots and areas. Item nodes 506 and 507 are nodes representing items related to the episode, such as historical or legendary cultural assets and works. Event node 508 is a node representing an event related to the episode, such as a historical or legendary phenomenon or incident.

[0081] [5. Other] The above-described embodiments can be used in appropriate combination. For example, the analysis unit 12 and the generation unit 13 of the information processing device 1 may perform processing using AI provided inside or outside the information processing device 1. The episode graph DB 1715 and the human-object-word graph DB 1716 may be provided in the storage unit 14. The storage unit 14 of the information processing device 1 may be provided in a server or the like that is communicably connected.

[0082] Furthermore, in the above embodiment, the information processing device 1 has been described as generating a two-dimensional episode graph and a two-dimensional human-monoword graph as an example, but it may also generate a three-dimensional episode graph and a three-dimensional human-monoword graph.

[0083] Furthermore, in the above embodiment, the information processing device 1 has been described as generating an episode graph using historical or other text, but this is not limiting. The information processing device 1 may generate an episode graph using, as episodes, text indicating usage situations, such as how content was used in the market. In such a case, an episode may be a text indicating what kind of person used what content at what event or location. In this way, an episode graph can serve as a tool or interface for organizing various causal relationships.

[0084] [6. Summary] (1, 7, 8) The information processing device 1 is an information processing device that generates a knowledge graph composed of a plurality of nodes and edges indicating the relationships between each of the plurality of nodes. This information processing device 1 includes a memory unit 14 that stores a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating the history included in the sentence, a first control unit 151 that controls output of the first knowledge graph from the memory unit 14 in a manner in which the episode nodes included in the first knowledge graph are connected by history edges, a reception unit 16 that receives instructions from a user, and a device that outputs information indicating episodes associated with the episode nodes included in the output first knowledge graph based on the instructions received by the reception unit 16. The knowledge graph includes a second control unit 152, and a third control unit 153 that controls the output of a second knowledge graph generated based on the first knowledge graph in a manner in which the two nodes are connected by an edge, the second knowledge graph including at least two nodes selected from a place node which is a node related to a location, a person node which is a node related to a person, an object node which is a node related to an object other than a person, and an event node which is a node related to an event, and an edge connecting the two nodes, wherein the at least two nodes are generated from an episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from a history edge.

[0085] Generally, people like and understand stories. This is because nerve cells in the human brain called "mirror neurons" allow people to imagine the scenes from the story when processing information from the story, and give people the feeling that they are in that scene.

[0086] Furthermore, while a typical knowledge graph shows how information is related and combined, as the amount of information increases it becomes more complex, making it difficult for people to read through the information.

[0087] The information processing device 1 generates a first knowledge graph (episode graph) using nodes (episode nodes) that indicate episodes contained in a text, such as history. Based on the first knowledge graph (episode graph), the information processing device 1 generates a second knowledge graph (person-thing-word graph) using four types of nodes: person nodes, thing nodes, event nodes, and place nodes. In this way, the information processing device 1 represents the multiple episodes contained in the text in a dual-structure knowledge graph.

[0088] In the first knowledge graph (episode graph), multiple episodes contained in a sentence are represented as episode nodes, and the history between episodes is connected by history edges. This allows the information (episode nodes) in the first knowledge graph (episode graph) to have a narrative quality, making it easier for people to read through the information (episode nodes) and understand the information in the first knowledge graph (episode graph). Furthermore, reading through the information organized in the first knowledge graph (episode graph) can lead to the discovery of new knowledge.

[0089] In the second knowledge graph (human-thing-word graph), information related to people, things, events, or places extracted from the episode nodes of the first knowledge graph is represented by four types of nodes: person nodes, object nodes, event nodes, and place nodes, and the history between nodes is connected by edges. This categorizes the information in the second knowledge graph (human-thing-word graph) into four types of nodes, allowing people to predict the type of content of the nodes (information), making it easier for people to read through the nodes (information) and understand the information in the second knowledge graph (human-thing-word graph). Furthermore, by reading through the information organized in the second knowledge graph (human-thing-word graph), people can discover new knowledge. Furthermore, categorizing the information in the second knowledge graph (human-thing-word graph) into four types of nodes allows for a game-like element, allowing people to select any node of interest from among person nodes, object nodes, event nodes, or place nodes. This allows people to enjoy reading through episodes that will provide new knowledge and acquire knowledge in a fun way. In this way, the information processing device 1 can generate a knowledge graph that is easy for people to see and understand, and that leads to the discovery of new knowledge.

[0090] (2) In such an information processing device 1, the process edges can include at least one of a physical process edge that indicates a process based on the passage of time or a procedure, and a logical process edge that indicates a process based on a causal relationship. This allows a person to visually understand that there is a physical process relationship and a logical process relationship between episode nodes connected by the process edges.

[0091] (3) The system may further include a generation unit 13 that analyzes text using the 5W1H method, treats sentences containing 4W elements as episodes, and treats sentences containing 1W1H elements as contexts, and generates episode nodes and context edges. This allows episodes, which are information (episode nodes) in the first knowledge graph (episode graph), to have a 5W1H narrative, making it easier for people to read through the episode nodes and understand the information in the first knowledge graph (episode graph).

[0092] (4) Here, the sentence may indicate a story related to history or legend, the person node may be a node representing the entity that is the subject of the episode, the place node may be a node representing a location related to the episode, the thing node may be a node representing an object related to the episode, and the event node may be a node representing an event related to the episode.

[0093] For example, stories about history or legends are a collection of anecdotes containing elements of the 5W1H system that relate to numerous events that occurred in the past, and can be considered knowledge that is liked and understood by many people. However, as history learned in school is often ridiculed as a mere "memorization subject," some people do not find history interesting as a story and see it as nothing more than a vast mass of information, leading to so-called "history haters."

[0094] The information processing device 1 classifies episodes of history, etc., represented in the second knowledge graph (human-thing-word graph) into the above-mentioned person nodes, thing nodes, event nodes, and place nodes, and can display the contents of the episodes in an organized state. This makes it easier for people to read episodes of history, etc., and can make them feel the interest of history.

[0095] (5) Furthermore, the system can further include a fourth control unit 154 that, when the receiving unit receives the designation of a specific node among the nodes included in the second knowledge graph (human-thing-word graph), outputs an explanation of the episode associated with the specific node and an explanation of the episode showing the relationship between the specific node and other nodes connected by edges. This outputs the explanation (content) of the episode associated with the node designated by the person and the content of the episode related to the designated node, so that by reading through the related episodes, the user can not only discover new knowledge but also acquire knowledge in a fun way.

[0096] (6) Furthermore, an episode includes time information when an event occurred, and the fourth control unit 154 can further arrange and display the multiple episodes in chronological order based on the time information when multiple episodes are associated with a specific node. In this way, when multiple episodes are associated with a specified node and these episodes include time information, the multiple episodes are arranged and displayed in chronological order. This makes it easier for people to read through the episodes and can enable them to acquire knowledge in a fun way. [Explanation of symbols]

[0097] 1. Information processing equipment 2. User terminal 11 Acquisition Department 12 Analysis Department 13 Generation part 14 Storage section 15 Control Unit 16 Reception Department 21 Display section 22 Terminal Acquisition Department 23 Instruction transmission unit 151 First Control Section 152 Second Control Section 153 Third Control Section 154 4th Control Section 170 CPU 171 memory 172 Storage Devices 173 connection port 174 Communication Circuits 175 Bus Line 301, 302, 303, 304, 305, 306, 307 Episode nodes 401, 402, 403, 404, 405 Longitudinal edges 501, 502, 503 nodes 504, 505 Field nodes 506, 507 Item node 508 Things Node 601, 602, 603, 604, 605, 606, 607, 608 Edge 1711 OS 1712 Browser Program 1713 Information processing device program 1715 Episode Graph DB 1716 Human and Object Word Graph Database 1717 User Terminal Program

Claims

1. An information processing device that generates a knowledge graph composed of a plurality of nodes and edges that indicate relationships with each of the plurality of nodes, a storage unit that stores a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating history included in the sentence; a first control unit that controls outputting the first knowledge graph from the storage unit in a manner in which the episode nodes included in the first knowledge graph are connected by the history edges; a reception unit that receives instructions from a user; a second control unit that outputs information indicating an episode associated with the episode node included in the output first knowledge graph based on the instruction received by the reception unit; a third control unit configured to output a second knowledge graph generated based on the first knowledge graph, the second knowledge graph including at least two nodes selected from a place node that is a node related to a location, a person node that is a node related to a person, an object node that is a node related to an object other than a person, and an event node that is a node related to an event, and an edge connecting the two nodes, the second knowledge graph being generated based on the first knowledge graph in a manner in which the at least two nodes are connected by the edge; the at least two nodes are generated from the episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from the historical edge; Information processing device.

2. The process edge includes at least one of a physical process edge indicating a process based on a time course or a procedure, and a logical process edge indicating a process based on a causal relationship. The information processing device according to claim 1 .

3. The system further includes a generation unit that performs a 5W1H analysis of the sentence, defines a sentence including each element of 4W (When, Where, Who, What) as an episode, and defines a sentence including each element of 1W1H (Why, How) as a history, and generates the episode node and the history edge. The information processing device according to claim 1 .

4. The text depicts a historical or legendary story, The person node is a node representing an entity that is the subject of the episode, the location node is a node representing a location related to the episode, the entity node is a node representing an entity related to the episode, The previous event node is a node representing an event related to the episode. The information processing device according to claim 1 .

5. a fourth control unit that, when the receiving unit receives a designation of a predetermined node among the nodes included in the second knowledge graph, outputs a description of an episode associated with the predetermined node and a description of an episode indicating a relationship between the predetermined node and other nodes connected by the edges; The information processing device according to claim 4 .

6. The episode includes time information when the event occurred; the fourth control unit further displays the plurality of episodes in chronological order based on the time information when the plurality of episodes are associated with the predetermined node. The information processing device according to claim 5 .

7. A program for generating a knowledge graph consisting of a plurality of nodes and edges indicating relationships between the plurality of nodes, a first control step of controlling output of a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating history included in the sentence, from a storage unit that stores the first knowledge graph in a manner in which the episode nodes included in the first knowledge graph are connected by the history edges; a receiving step of receiving an instruction from a user; a second control step of outputting information indicating the episode associated with the episode node included in the output first knowledge graph based on the instruction received in the receiving step; a third control step of controlling a second knowledge graph generated based on the first knowledge graph to include at least two nodes selected from a place node that is a node related to a location, a person node that is a node related to a person, an object node that is a node related to an object other than a person, and an event node that is a node related to an event, and an edge connecting the two nodes, in a manner in which the two nodes are connected by the edge; the at least two nodes are generated from the episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from the historical edge; program.

8. 1. A computer-implemented method for generating a knowledge graph comprising a plurality of nodes and edges representing relationships between the plurality of nodes, the method comprising: a first control step of controlling output of a first knowledge graph generated by analyzing a sentence, the first knowledge graph including episode nodes indicating each of a plurality of episodes included in the sentence and history edges indicating history included in the sentence, from a storage unit that stores the first knowledge graph in a manner in which the episode nodes included in the first knowledge graph are connected by the history edges; a receiving step of receiving an instruction from a user; a second control step of outputting information indicating the episode associated with the episode node included in the output first knowledge graph based on the instruction received in the receiving step; a third control step of controlling output of a second knowledge graph generated based on the first knowledge graph, the second knowledge graph including at least two nodes selected from a place node which is a node related to a location, a person node which is a node related to a person, an object node which is a node related to an object other than a person, and an event node which is a node related to an event, and an edge connecting the two nodes, in a manner in which the two nodes are connected by the edge; the at least two nodes are generated from the episode node included in the first knowledge graph, and the edge connecting the two nodes is generated from the historical edge; A computer-implemented method for generating knowledge graphs.

Images