Information processing device, information processing system, information processing method, and recording medium

Abstract

The present invention provides users with a description of a relationship between a plurality of elements. An information processing device according to the present invention includes an acquisition unit, an input unit, and an output unit. The acquisition unit acquires information indicating the strength of the relationship between the plurality of elements, the input unit inputs information indicating the strength of the relationship between the plurality of elements to a generation model, and the output unit outputs information expressing the relationship between the plurality of elements in language on the basis of the result of analysis by the generation model.

Description

Information processing device, information processing system, information processing method, and recording medium

[0001] The present disclosure relates to an information processing device, an information processing system, an information processing method, and a recording medium, and more particularly to an information processing device, an information processing system, an information processing method, and a recording medium that provide a user with an analysis result of information.

[0002] In technical fields such as behavior prediction and anomaly detection, related techniques for analyzing causal relationships between elements are known.

[0003] Patent Literature 1 describes a method for presenting causal elements, resulting elements, and the strength of causality in a table format. A related technology described in Patent Literature 1 provides a causal graph in which the causal elements and resulting elements are represented by nodes, and the strength of causality is represented by edges connecting the nodes.

[0004] Japanese Patent Application Laid-Open No. 2022-15336

[0005] In related technologies, information representing the strength of the relationships between multiple elements is provided in the form of numerical data or graphical data, but as the number of elements increases, the numerical data and graphical data become more complex, which creates a problem in that the information provided by related technologies is difficult for users to understand.

[0006] The present disclosure has been made in view of the above-mentioned problems, and its purpose is to provide a user with an explanation of the relationships between multiple elements.

[0007] An information processing device according to one aspect of the present disclosure includes an acquisition means for acquiring information representing the strength of the relationship between multiple elements, an input means for inputting the information representing the strength of the relationship between the multiple elements into a generative model, and an output means for outputting information expressing the relationship between the multiple elements in language based on the analysis results by the generative model.

[0008] An information processing system according to one aspect of the present disclosure comprises an information processing device, one or more user terminals used by a user to input information or instructions to the information processing device, and an information analysis execution device trained to contextualize the input data and output output data according to the context of the input data, wherein the information processing device acquires information representing the strength of the relationship between multiple elements from the one or more user terminals, inputs the information representing the strength of the relationship between the multiple elements to the information analysis execution device, and outputs information expressing the relationship between the multiple elements in language based on the analysis results by the information analysis execution device.

[0009] In an information processing method according to one aspect of the present disclosure, a computer acquires information representing the strength of the relationships between multiple elements, inputs the information representing the strength of the relationships between the multiple elements into a generative model, and outputs information expressing the relationships between the multiple elements in language based on the analysis results of the generative model.

[0010] A recording medium according to one aspect of the present disclosure stores a program for causing a computer to execute the following processes: acquiring information representing the strength of the relationship between multiple elements; inputting the information representing the strength of the relationship between the multiple elements into a generative model; and outputting information expressing the relationship between the multiple elements in language based on the analysis results of the generative model.

[0011] According to one aspect of the present disclosure, an explanation of the relationships between multiple elements can be provided to the user.

[0012] 1 is a conceptual diagram showing an example of the operation of a system including an information processing device according to an embodiment. FIG. 2 is a diagram showing an example of input data (information representing the strength of relationships between multiple elements) input to a generative model in a system including an information processing device according to an embodiment. FIG. 3 is a diagram showing an example of output data (information based on the analysis results by the generative model) output from the system. FIG. 4 is a block diagram showing the configuration of an information processing device according to an embodiment. FIG. 5 is a flowchart showing the operation of an information processing device according to an embodiment. FIG. 6 is a conceptual diagram showing an example of the operation of a system including an information processing device according to an embodiment. FIG. 7 is a block diagram showing the configuration of an information processing device according to an embodiment. FIG. 8 is a flowchart showing the operation of an information processing device according to an embodiment. FIG. 9 is a conceptual diagram showing an example of the operation of a system including an information processing device according to an embodiment. FIG. 10 is a diagram showing an example of a causal graph, which is information representing the strength of relationships between multiple elements. FIG. 11 is a diagram showing an example of output data (discussion screen) output by an information processing device according to an embodiment. FIG. 11 is a block diagram showing the configuration of an information processing device according to an embodiment. FIG. 12 is a flowchart showing the operation of an information processing device according to an embodiment. FIG. 13 is a diagram showing an example of an information processing system according to an embodiment.

[0013] Some embodiments of the present disclosure will be described with reference to the drawings.

[0014] First Embodiment A first embodiment of the present disclosure will be described with reference to FIGS. 1 to 5. FIG.

[0015] (Example of System) Fig. 1 is a conceptual diagram showing an example of the operation of the system according to embodiment 1. As shown in Fig. 1, information representing the strength of the relationship between a plurality of elements is input to a generative model.

[0016] For example, the information representing the strength of the relationships between the multiple elements is a correlation coefficient matrix composed of correlation coefficients representing the correlations between the multiple elements. In another example, the information representing the strength of the relationships between the multiple elements is a causal graph ( FIG. 10 ) representing the causal relationships between the multiple elements. The causal graph ( FIG. 10 ) is an example of chart data.

[0017] A generative model is an information analysis program that is trained, typically through machine learning or deep learning, to contextualize input data and output data that corresponds to the context of the input data. A generative model analyzes information that represents the strength of relationships between multiple elements and outputs the analysis results.

[0018] More specifically, a generative model is a model that learns the relationships between words in a sentence (including letters, symbols, and numbers) and the relationships between elements shown in a diagram, and generates related strings related to a target string from the target string. By using a generative model that has been trained on sentences and paragraphs from various contexts, it is possible to generate related strings with appropriate content related to the target string.

[0019] For example, a case where a generative model is used in question answering will be described. The generative model receives an input question such as "What kind of country is Japan?" as a target string. The generative model generates a string such as "Japan is an island country in the Northern Hemisphere..." as an answer to the question.

[0020] The method of training the generative model is not particularly limited, but as an example, the generative model may be trained to output at least one sentence that includes an input string.

[0021] As a specific example, a generative model is a Generative Pre-trained Transformer (GPT) that outputs a sentence including an input string by predicting a string that is likely to follow the input string. Other generative models include a Text-to-Text Transfer Transformer (T5), a Bidirectional Encoder Representations from Transformers (BERT), a Robustly optimized BERT approach (RoBERTa), and an Efficiently Learning an Encoder that Classifies Token Replacements Accurately (ELECTRA).

[0022] Furthermore, the character strings generated by the generative model are not limited to natural languages. For example, the generative model may output an artificial language (such as program source code) in response to a character string input in a natural language. For example, the generative model may receive an input question such as, "How do I retrieve data containing a specific character string from a database?" as a target character string. The generative model may output program source code for performing database processing. Alternatively, the generative model may output a natural language corresponding to a character string input in an artificial language.

[0023] Furthermore, the content generated by a generative model is not limited to character strings. For example, the generative model may generate image data, video data, audio data, or other data formats corresponding to the input data.

[0024] An information processing device 10 (FIG. 4) described below creates information based on the analysis results of the input data, based on the analysis results output by the generative model.

[0025] As shown in FIG. 1, information based on the analysis results of the generative model is output.

[0026] For example, the output data may be a comment on the relationships between the elements. In another example, the output data may be at least one of facts, observations, and suggestions (e.g., policy proposals) derived from the information representing the strength of the relationships between the elements.

[0027] (Example of Input Data) Figure 2 shows an example of input data input to the generative model. In the example shown in Figure 2, the strength of the relationship between multiple elements (e.g., "livability" and "medical and nursing care system") is expressed as a numerical value. Specifically, these numerical values ​​are correlation coefficients between one element (variable) and another element.

[0028] (Example of Output Data) FIG. 3 shows an example of output data created based on the analysis results of the generative model. In the example shown in FIG. 3, comments about the relationship between multiple elements are shown. For example, the example shown in FIG. 3 states that "the size of a house itself is not likely to be directly related to the motivation for purchasing a house." Specifically, this comment explains (considers) the magnitude of the influence that one element (variable) called "house size" has on another element called "motivation for purchasing a house."

[0029] (Configuration of Information Processing Apparatus 10) The configuration of the information processing apparatus 10 according to the first embodiment will be described with reference to Fig. 4. Fig. 4 is a block diagram showing the configuration of the information processing apparatus 10.

[0030] As shown in FIG. 4, the information processing device 10 includes an acquisition unit 11, an input unit 12, and an output unit 13.

[0031] The acquisition unit 11 acquires information representing the strength of the relationship between a plurality of elements. The acquisition unit 11 is an example of an acquisition means. The information representing the strength of the relationship between a plurality of elements may be numerical data such as a vector or a matrix, or may be diagram data such as a graph or a diagram.

[0032] For example, the acquisition unit 11 acquires the results of a causal analysis performed on information about a plurality of elements from a causal analysis device (not shown). The results of the causal analysis indicate whether or not one or more other elements (causes) have an effect on one element (result), and the magnitude of the effect. In this example, the results of the causal analysis performed by the causal analysis device are information representing the strength of the causal relationship between the plurality of elements (for example, a coefficient or matrix representing the strength of the causal relationship between the plurality of elements).

[0033] Alternatively, the acquisition unit 11 may acquire information indicating the strength of the relationship between a plurality of elements, which information is input by the user to the information processing device 10 (second embodiment).

[0034] The acquisition unit 11 outputs information indicating the strength of the relationship between a plurality of elements to the input unit 12 .

[0035] The input unit 12 inputs information representing the strength of relationships between multiple elements to the generative model. The input unit 12 is an example of an input means.

[0036] For example, the input unit 12 receives information representing the strength of relationships between multiple elements from the acquisition unit 11. The input unit 12 inputs the information representing the strength of relationships between multiple elements to the generative model ( FIG. 1 ). In one example, the input unit 12 inputs the information representing the strength of relationships between multiple elements to the generative model in the form of numerical values ​​such as vectors or matrices. In another example, the input unit 12 inputs the information representing the strength of relationships between multiple elements to the generative model in the form of images such as graphs or diagrams.

[0037] The input unit 12 then instructs the generative model to analyze the input information. At this time, the input unit 12 may accept input of a prompt (instruction statement) from the user. The generative model analyzes information indicating the strength of relationships between multiple elements in accordance with the instructions from the input unit 12 and the prompt, and outputs the analysis results.

[0038] The input unit 12 notifies the output unit 13 that the generative model has executed the analysis of the information.

[0039] The output unit 13 outputs information that expresses in language the relationship between one or more specified elements based on the analysis results (for example, facts, considerations, and conclusions) obtained by the generative model.

[0040] For example, after receiving a notification from the input unit 12, the output unit 13 acquires, from the generative model, an analysis result of information indicating the strength of relationships between multiple elements. The output unit 13 outputs information based on the acquired analysis result.

[0041] In one example, the output unit 13 displays, on a user terminal (not shown), a discussion screen ( FIG. 11 ) on which the user can view the analysis results of the generative model. The discussion screen may include input data (e.g., information indicating the strength of the relationships between multiple elements) along with the analysis results of the generative model. The discussion screen may also include comments entered by the user (e.g., the user's opinion on the analysis results of the generative model).

[0042] (Operation of Information Processing Apparatus 10) The operation of the information processing apparatus 10 according to the first embodiment will be described with reference to Fig. 5. Fig. 5 is a flowchart showing the operation of the information processing apparatus 10.

[0043] As shown in FIG. 5, first, the acquisition unit 11 acquires information indicating the strength of the relationship between a plurality of elements (S101).

[0044] The acquisition unit 11 outputs information indicating the strength of the relationship between a plurality of elements to the input unit 12 .

[0045] Next, the input unit 12 inputs information representing the strength of the relationships between a plurality of elements to the generative model (FIG. 1) (S102).

[0046] The input unit 12 notifies the output unit 13 that the generative model has executed the analysis of the information.

[0047] Next, the output unit 13 outputs information that expresses the relationships between multiple elements in language based on the analysis results using the generative model (S103).

[0048] For example, the output unit 13 displays, on a user terminal (not shown), a consideration screen (FIG. 11) on which the user can view the analysis results based on the generative model.

[0049] This completes the operation of the information processing device 10 according to the first embodiment.

[0050] (Effects of this embodiment) According to the configuration of this embodiment, the acquisition unit 11 acquires information representing the strength of the relationships between multiple elements. The input unit 12 inputs the information representing the strength of the relationships between multiple elements to a generative model. The output unit 13 outputs information expressing the relationships between the multiple elements in language based on the analysis results of the generative model.

[0051] In this way, information that expresses the relationship between multiple elements in language is output, rather than information that directly represents the strength of the relationship between multiple elements, so an explanation of the relationship between multiple elements can be provided to the user.

[0052] [Second Embodiment] A second embodiment will be described with reference to Fig. 6 to Fig. 8. In the second embodiment, a configuration will be described in which a user is allowed to specify (select) information to be analyzed by a generative model.

[0053] In the second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals, and the description thereof will be omitted.

[0054] (Example of System) Fig. 6 is a conceptual diagram showing an example of the operation of the system according to embodiment 2. As shown in Fig. 6, in the system, information representing the strength of the relationship between a plurality of elements is input to a generative model (S1).

[0055] For example, the information representing the strength of the relationships between the multiple elements is a correlation coefficient matrix composed of correlation coefficients representing the correlations between the multiple elements. In another example, the information representing the strength of the relationships between the multiple elements is a causal graph ( FIG. 10 ) representing the causal relationships between the multiple elements.

[0056] The generative model analyzes information indicating the strength of relationships between multiple elements in accordance with instructions from the input unit 12 and prompts, and outputs the analysis results (S2).

[0057] An information processing device 20 (FIG. 7) described below creates information based on the analysis results of the input data, based on the analysis results generated by the generative model.

[0058] The system outputs information based on the analysis results of the generative model (S3).

[0059] For example, the output data may be comments about the relationships between the elements (FIG. 3). In one example, the comments may include at least one of facts, observations, and suggestions (e.g., policy proposals) derived from the relationships between the elements.

[0060] As shown in FIG. 6, the user inputs instructions to the user terminal 100 to select which relationships between multiple elements should be reanalyzed by the generative model (e.g., selecting elements A and B, and selecting elements E and F) (S4).

[0061] An information processing device 20 (FIG. 7) described below inputs information representing the strength of the relationship between the elements selected by the user into the generative model in accordance with instructions from the user terminal 100 (S5).

[0062] In one example, the information representing the strength of the relationships between the elements selected by the user is a correlation coefficient matrix that includes only the elements (correlation coefficients) designated for reanalysis by the generative model, or a correlation coefficient matrix in which all elements not designated for reanalysis by the generative model are set to 0.

[0063] Again, the generative model analyzes the information indicating the strength of the relationships between multiple elements in accordance with instructions from the input unit 12 and prompts, and outputs the analysis results (S2 for the second time).

[0064] (Configuration of Information Processing Device 20) The configuration of the information processing device 20 according to the second embodiment will be described with reference to Fig. 7. Fig. 7 is a block diagram showing the configuration of the information processing device 20.

[0065] 7, the information processing device 20 includes an acquisition unit 11, an input unit 12, and an output unit 13. The information processing device 20 further includes a reception unit 24.

[0066] The receiving unit 24 receives a user operation for selecting which relationships between a plurality of elements are to be analyzed by the generative model. The receiving unit 24 is an example of a receiving means.

[0067] For example, the receiving unit 24 receives a user operation for selecting elements to be analyzed by the generative model from the user terminal 100 ( FIG. 6 ). For example, the user uses the user terminal 100 to select elements A and B, and elements E and F, to be analyzed by the generative model.

[0068] The reception unit 24 receives from the user terminal 100 instructions input to the user terminal 100 (for example, selection of elements A and B, and selection of elements E and F).

[0069] The receiving unit 24 outputs information indicating the instruction input to the user terminal 100 to the acquiring unit 11 .

[0070] The acquiring unit 11 acquires information indicating the strength of the relationship between a plurality of elements. In the second embodiment, the acquiring unit 11 receives information indicating an instruction input to the user terminal 100 from the accepting unit 24.

[0071] The acquisition unit 11 acquires information representing the strength of the causal relationship between the elements selected by the user (for example, a coefficient or matrix representing the strength of the causal relationship between the elements selected by the user) from a causal analysis device not shown in accordance with instructions input to the user terminal 100.

[0072] The acquisition unit 11 outputs, to the input unit 12, information indicating the strength of the causal relationship between the elements selected by the user.

[0073] As in the first embodiment, the input unit 12 inputs information (e.g., a correlation coefficient matrix) representing the strength of the causal relationships between multiple elements to the generative model. However, in the second embodiment, the input unit 12 inputs information representing the strength of the causal relationships between elements selected by the user to the generative model. As a result, only the multiple elements selected by the user become the targets of analysis by the generative model.

[0074] The input unit 12 notifies the output unit 13 that the generative model has executed the analysis of the information.

[0075] The output unit 13 receives the analysis results (e.g., facts, considerations, and conclusions) from the generative model, and outputs information based on the analysis results from the generative model (e.g., a consideration screen (FIG. 11)).

[0076] (Operation of Information Processing Device 20) The operation of the information processing device 20 according to the second embodiment will be described with reference to Fig. 8. Fig. 8 is a flowchart showing the operation of the information processing device 20.

[0077] 8, first, the acquiring unit 11 acquires information representing the strength of the relationship between a plurality of elements (S201). The information representing the strength of the relationship between a plurality of elements is, for example, a correlation coefficient indicating the correlation between the plurality of elements or a coefficient representing the strength of the causal relationship between the plurality of elements.

[0078] The acquisition unit 11 outputs information indicating the strength of the relationship between a plurality of elements to the input unit 12 .

[0079] Next, the input unit 12 inputs information representing the strength of the relationships between a plurality of elements to the generative model (FIG. 6) (S202).

[0080] The input unit 12 notifies the output unit 13 that the generative model has executed the analysis of the information.

[0081] Next, the output unit 13 outputs information expressing the relationships between the multiple elements in language based on the analysis results using the generative model (S203). For example, the output unit 13 displays, on the user terminal 100 (FIG. 6), a discussion screen (FIG. 11) on which the user can view the analysis results using the generative model.

[0082] The receiving unit 24 receives a user operation for selecting which relationships among a plurality of elements are to be analyzed by the generative model (S204).

[0083] If a user operation for selecting which relationships among a plurality of elements should be analyzed by the generative model is accepted (Yes in S205), the flow returns to step S202.

[0084] On the other hand, if the user operation is not accepted (No in S205), the operation of the information processing apparatus 20 according to the second embodiment ends.

[0085] (Effects of this embodiment) According to the configuration of this embodiment, the acquisition unit 11 acquires information representing the strength of the relationships between multiple elements. The input unit 12 inputs the information representing the strength of the relationships between multiple elements to a generative model. The output unit 13 outputs information expressing the relationships between the multiple elements in language based on the analysis results of the generative model.

[0086] In this way, information that expresses the relationship between multiple elements in language is output, rather than information that directly represents the strength of the relationship between multiple elements, so an explanation of the relationship between multiple elements can be provided to the user.

[0087] Furthermore, according to the configuration of this embodiment, the receiving unit 24 receives a user operation for selecting which relationships among multiple elements should be analyzed by the generative model. The input unit 12 inputs the information selected by the user to the generative model. The output unit 13 outputs information expressing in language the relationships between the elements selected by the user based on the analysis results by the generative model.

[0088] This allows only the information that the user needs to be presented in a format that is easy for the user to understand.

[0089] 9 to 13, a third embodiment of the present disclosure will be described. In the third embodiment, a configuration will be described in which information indicating the strength of the relationship between a plurality of elements is created from information about the plurality of elements.

[0090] (Example of System) Fig. 9 is a conceptual diagram showing an example of the operation of the system according to embodiment 3. As shown in Fig. 9, information on a plurality of elements (for example, questionnaire results on a plurality of elements) is input to the system.

[0091] The system performs a causal analysis on information about multiple elements, and as a result of the causal analysis, information indicating the strength of the causal relationships between multiple elements (for example, a causal graph ( FIG. 10 ) or a matrix representing the causal relationships ( FIG. 9 )) is obtained.

[0092] As shown in FIG. 9, information representing the strength of relationships between multiple elements is input into a generative model (generative AI) (e.g., a large-scale generative model).

[0093] For example, the input data may be a correlation coefficient matrix composed of correlation coefficients that represent correlations between multiple elements. In another example, the input data may be a causal graph ( FIG. 10 ) that represents causal relationships between multiple elements. An example of information representing the strength of relationships between multiple elements will be described later.

[0094] The generative model analyzes information that represents the strength of the relationships between multiple elements and outputs the analysis results.

[0095] An information processing device 30 (FIG. 12) described later creates information based on the analysis results output by the generative model.

[0096] As shown in FIG. 9, information based on the analysis results of the generative model is output.

[0097] For example, the output data may be comments about the relationships between the elements. In one example, the comments may include at least one of facts, observations, and suggestions (e.g., policy proposals) derived from the relationships between the elements. An example of information based on the analysis results of the generative model will be described later.

[0098] (Example of Input Data: Causal Graph) Fig. 10 is an example of a causal graph. As shown in Fig. 10, the causal graph is made up of a plurality of nodes and arrows connecting the nodes.

[0099] As shown in Figure 10, in a causal graph, one node is connected to one or more other nodes by arrows. Each node represents a result or a cause in a causal relationship. Arrows connect nodes in a causal relationship. The node at the end of an arrow is the result, and the node at the start of an arrow is the cause.

[0100] In this example, the node "Motivation for Purchasing a Home" is the final result, that is, the "Objective Variable." The nodes other than the node "Motivation for Purchasing a Home" are the "Explanatory Variables."

[0101] The numbers attached to the lines connecting nodes are correlation coefficients that represent the relationship between the nodes. If the value between nodes is "0," there is no relationship between those nodes. On the other hand, if the value between nodes is other than "0," there is a relationship between those nodes. The strength of the relationship between nodes is compared based on the magnitude of the numbers. For example, the node "Number of Children" has a relationship strength of "0.28" with the node "House Size." On the other hand, the node "Salary" has a relationship strength of "0.29" with the node "House Size." Therefore, it can be seen that "House Size" is more influenced by "Salary" than by "Number of Children." Note that solid arrows indicate a positive effect on "Motivation for Purchasing a Home." On the other hand, dashed arrows indicate a negative effect on "Motivation for Purchasing a Home." For example, the node "Number of Children" has a relationship strength of "0.28" with "Motivation for Purchasing a Home." On the other hand, the node "Price (of Real Estate)" has a relationship strength of "0.4" with "Motivation for Purchasing a Home."

[0102] The generative model (Figure 9) may be input with the causal graph itself, as in this example (Figure 10), or with a correlation coefficient matrix whose elements are the numerical values ​​(correlation coefficients) between nodes in the causal graph.

[0103] (Example of Output Data: Discussion Screen) Fig. 11 shows an example of information based on the analysis results using a generative model. This example is a discussion screen that allows the user to view the analysis results using a generative model. The discussion screen is displayed on the user terminal 100 (Fig. 14) or the like.

[0104] As shown in FIG. 11 , the consideration screen includes a causal graph (an example of information indicating the strength of relationships between multiple elements) and analysis results (e.g., facts, considerations, and conclusions) from the generative model. Furthermore, a comment field is also shown below these. In addition to inputting the analysis results from the generative model into the comment field, the user can also enter their own opinions in the comment field using the user terminal 100.

[0105] By viewing the consideration screen shown in Figure 11, the user can confirm the analysis results using the generative model, and can also directly confirm information representing the strength of the relationships between multiple elements from the causal graph.

[0106] (Configuration of Information Processing Device 30) The configuration of the information processing device 30 according to the third embodiment will be described with reference to Fig. 12. Fig. 12 is a block diagram showing the configuration of the information processing device 30.

[0107] As shown in FIG. 12, the information processing device 30 includes an acquisition unit 31 , an input unit 32 , and an output unit 33 .

[0108] The acquisition unit 31 performs a causal analysis on information relating to a plurality of elements.

[0109] For example, the acquisition unit 31 acquires information about a plurality of elements from information collected in advance in a database (not shown). The acquisition unit 31 performs a causal analysis on the information about the plurality of elements to acquire a coefficient representing the strength of the causal relationship between the plurality of elements.

[0110] The acquisition unit 31 outputs to the input unit 32 information indicating correlation coefficients between the plurality of elements, which information is obtained by performing a causal analysis on information relating to the plurality of elements.

[0111] The input unit 32 inputs a correlation coefficient matrix, whose elements are correlation coefficients between a plurality of elements, to the generative model. The input unit 32 is an example of an input means.

[0112] For example, the input unit 32 receives information indicating correlation coefficients between a plurality of elements from the acquisition unit 31. The input unit 32 creates a correlation coefficient matrix having the correlation coefficients between the plurality of elements as elements. Then, the input unit 32 inputs the created correlation coefficient matrix to the generative model.

[0113] The input unit 32 notifies the output unit 33 that the generative model has executed the analysis of the information.

[0114] The generative model analyzes information that represents the strength of the relationships between multiple elements and outputs the analysis results.

[0115] The output unit 33 outputs information based on the analysis result of the correlation coefficient matrix by the generative model. The output unit 33 is an example of an output means.

[0116] For example, after receiving a notification from the input unit 32, the output unit 33 acquires, from the generative model, an analysis result of information indicating the strength of relationships between multiple elements. The output unit 13 outputs information based on the acquired analysis result.

[0117] In one example, the output unit 33 displays, on the user terminal 100 (FIG. 9), a discussion screen (FIG. 11) on which the user can view the analysis results of the generative model. The discussion screen may include input data (e.g., information indicating the strength of the relationships between multiple elements) along with the analysis results of the generative model. Furthermore, the discussion screen may include comments entered by the user (e.g., opinions on the analysis results of the generative model).

[0118] (Operation of Information Processing Device 30) The operation of the information processing device 30 according to the third embodiment will be described with reference to Fig. 13. Fig. 13 is a flowchart showing the operation of the information processing device 30.

[0119] 13 , first, the acquisition unit 31 acquires information about a plurality of elements. The acquisition unit 31 performs a causal analysis on the information about the plurality of elements (S301). As a result, the acquisition unit 31 acquires information indicating the strength of the causal relationship between the plurality of elements.

[0120] The acquisition unit 31 outputs information indicating the strength of the causal relationship between a plurality of elements to the input unit 32 .

[0121] Next, the input unit 32 inputs a matrix representing the causal relationships between a plurality of elements into the generative model (FIG. 9) (S302).

[0122] The input unit 32 notifies the output unit 33 that the generative model has executed the analysis of the information.

[0123] Next, the output unit 33 outputs information expressing the causal relationships among the elements in language based on the analysis results of the matrix using the generative model (S303). For example, the output unit 13 displays, on the user terminal 100 (FIG. 9), a discussion screen (FIG. 11) on which the user can view the analysis results using the generative model.

[0124] This completes the operation of the information processing device 30 according to the third embodiment.

[0125] (Effects of this embodiment) According to the configuration of this embodiment, the acquisition unit 31 acquires information indicating the strength of the causal relationships between the multiple elements, obtained by performing causal analysis on information about the multiple elements. The input unit 32 inputs a matrix indicating the causal relationships between the multiple elements to a generative model. The output unit 33 outputs information based on the analysis result of the matrix by the generative model.

[0126] In this way, information that expresses the causal relationships between multiple elements in language is output, rather than information that directly represents the strength of the relationships between multiple elements (here, a matrix that represents the causal relationships between multiple elements), so it is possible to provide the user with an explanation of the causal relationships between multiple elements.

[0127] [Fourth Embodiment] A fourth embodiment will be described with reference to Fig. 14. In this fourth embodiment, an example of the configuration of an information processing system including any one of the information processing devices 10, 20, and 30 described in the first to third embodiments will be described.

[0128] (Information Processing System 1) Fig. 14 is a diagram schematically illustrating an example of the configuration of an information processing system 1 according to the fourth embodiment. As shown in Fig. 14, the information processing system 1 includes an information processing device 10 (20, 30), a user terminal 100, and an information analysis execution device 200. Note that the number of user terminals 100 included in the information processing system 1 is not limited to three.

[0129] Here, "information processing device 10 (20, 30)" means "any of the information processing devices 10, 20, 30 according to the first to third embodiments."

[0130] The user terminal 100 and the information processing device 10 (20, 30) are communicatively connected. The user terminal 100 is used by a user to input information (data) and instructions to the information processing device 10 (20, 30). For example, the user inputs information about multiple elements or information indicating the strength of the relationship between multiple elements to the user terminal 100. The user also inputs instructions (prompts) for the information analysis execution device 200 to the user terminal 100.

[0131] The information processing device 10 (20, 30) acquires information representing the strength of the relationship between a plurality of elements. As described in the first and second embodiments, the information processing device 10 (20, 30) may acquire the information itself representing the strength of the relationship between a plurality of elements. Alternatively, as described in the third embodiment, the information processing device 10 (20, 30) may acquire information regarding a plurality of elements, and then perform causal analysis on the information regarding the plurality of elements to acquire a coefficient representing the strength of the causal relationship between the plurality of elements.

[0132] The information processing device 10 (20, 30) inputs information representing the strength of the relationships between multiple elements to the information analysis execution device 200. At this time, the information processing device 10 (20, 30) may convert the information representing the strength of the relationships between multiple elements into a format compatible with the information analysis execution device 200, and then input the information to the information analysis execution device 200.

[0133] The information analysis execution device 200 is pre-trained to contextualize input data (e.g., text data, image data) and output output data according to the context of the input data. The information analysis execution device 200 is, for example, a text generation model (e.g., a large-scale generative model) that analyzes text data, or a generative model that analyzes input chart data.

[0134] The information analysis execution device 200 analyzes information that indicates the strength of the relationships between multiple elements, which information is input from the information processing device 10 (20, 30), and outputs the analysis results.

[0135] The information processing device 10 (20, 30) acquires the analysis result of information indicating the strength of the relationship between multiple elements from the information analysis execution device 200. Then, the information processing device 10 (20, 30) outputs information that expresses the relationship between the multiple elements in language based on the analysis result by the information analysis execution device 200.

[0136] (Effects of this embodiment) According to the configuration of this embodiment, the information processing device 10 (20, 30) acquires information representing the strength of the relationships between multiple elements. The information processing device 10 (20, 30) inputs the information representing the strength of the relationships between multiple elements to the information analysis execution device 200. The information processing device 10 (20, 30) outputs information that expresses the relationships between the multiple elements in language, based on the analysis results by the information analysis execution device 200.

[0137] In this way, information that expresses the relationship between multiple elements in language is output, rather than information that directly represents the strength of the relationship between multiple elements (here, the correlation coefficient matrix), so that an explanation of the relationship between multiple elements can be provided to the user.

[0138] (Hardware Configuration) Each of the components of the information processing devices 10, 20, and 30 described in the first to third embodiments represents a functional block. Some or all of these components are realized by an information processing device such as that shown in Fig. 15. Fig. 15 is a block diagram showing an example of the hardware configuration of the information processing device.

[0139] 15 , the computer 110 includes a CPU (Central Processing Unit) 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. These components are connected to each other via a bus 121 so as to be able to communicate data with each other. Note that the computer 110 may include a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array) in addition to or instead of the CPU 111.

[0140] The CPU 111 loads the program (code) of this embodiment stored in the storage device 113 into the main memory 112 and executes it in a predetermined order to perform various calculations. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). The program of this embodiment is provided in a state stored in a computer-readable recording medium 120. The program of this embodiment may be distributed over the Internet connected via the communication interface 117.

[0141] Specific examples of the storage device 113 include a hard disk drive and a semiconductor storage device such as a flash memory. The input interface 114 mediates data transmission between the CPU 111 and input devices 118 such as a keyboard and a mouse. The display controller 115 is connected to a display device 119 and controls the display on the display device 119.

[0142] The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, reads programs from the recording medium 120, and writes processing results from the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and other computers.

[0143] Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash (registered trademark)) and SD (Secure Digital), magnetic recording media such as flexible disks, or optical recording media such as CD-ROMs (Compact Disk Read Only Memory).

[0144] (Supplementary Notes) A ​​part or all of the above-described embodiments can be described as, but are not limited to, the following supplementary notes.

[0145] (Supplementary Note 1) An information processing device comprising: an acquisition means for acquiring information representing the strength of the relationship between a plurality of elements; an input means for inputting the information representing the strength of the relationship between the plurality of elements into a generative model; and an output means for outputting information expressing the relationship between the plurality of elements in language based on the analysis results by the generative model.

[0146] (Supplementary Note 2) The information processing device according to Supplementary Note 1, wherein the information representing the strength of the relationship between the plurality of elements is numerical data or chart data representing the strength of the relationship between the plurality of elements.

[0147] (Supplementary Note 3) The information processing device according to Supplementary Note 1, wherein the plurality of elements include a target variable that is a result and a plurality of explanatory variables that are causes.

[0148] (Supplementary Note 4) The information processing device according to Supplementary Note 1, wherein the generative model is an information analysis program that is trained to contextualize input data and output output data according to the context of the input data.

[0149] (Supplementary Note 5) The information processing device described in any one of Supplementary Notes 1 to 4, characterized in that the acquisition means acquires the information representing the strength of the relationship between the plurality of elements, obtained by performing causal analysis on information regarding the plurality of elements.

[0150] (Supplementary Note 6) The information processing device described in any one of Supplementary Notes 1 to 4, characterized in that the input means inputs the information representing the strength of the relationship between the multiple elements to the generative model in the form of a matrix whose elements are numerical values ​​representing the strength of the causal relationship between the multiple elements.

[0151] (Supplementary Note 7) The information processing device described in any one of Supplementary Notes 1 to 4, characterized in that the output means outputs comments about the relationship between the multiple elements as the information based on the analysis results of the information representing the strength of the relationship between the multiple elements by the generative model.

[0152] (Supplementary Note 8) The information processing device described in any one of Supplementary Notes 1 to 4, further comprising a receiving means for receiving a user operation to select the information representing the strength of the relationship between the multiple elements to be input into the generative model.

[0153] (Supplementary Note 9) An information processing method in which a computer acquires information representing the strength of the relationships between multiple elements, inputs the information representing the strength of the relationships between the multiple elements into a generative model, and outputs information expressing the relationships between the multiple elements in language based on the analysis results by the generative model.

[0154] (Supplementary Note 10) The information processing method according to Supplementary Note 9, wherein the information representing the strength of the relationship between the plurality of elements is numerical data representing the strength of the relationship between the plurality of elements.

[0155] (Supplementary Note 11) The information processing method according to Supplementary Note 9, wherein the plurality of elements includes a target variable that is a result and a plurality of explanatory variables that are causes.

[0156] (Supplementary Note 12) The information processing method according to Supplementary Note 9, wherein the generative model is an information analysis program trained to contextualize input data and output output data according to the context of the input data.

[0157] (Supplementary Note 13) The information processing method described in any one of Supplementary Notes 9 to 12, characterized in that the computer obtains the information representing the strength of the relationship between the multiple elements by performing causal analysis on information regarding the multiple elements.

[0158] (Supplementary Note 14) The information processing method described in any one of Supplementary Notes 9 to 12, characterized in that the computer inputs the information representing the strength of the relationship between the multiple elements to the generative model in the form of a matrix whose elements are numbers representing the strength of the causal relationship between the multiple elements.

[0159] (Appendix 15) An information processing method described in any one of Appendices 9 to 12, characterized in that the computer outputs comments about the relationship between the multiple elements as the information based on the analysis results of the information representing the strength of the relationship between the multiple elements by the generative model.

[0160] (Supplementary Note 16) The information processing method described in any one of Supplementary Notes 9 to 12, characterized in that the computer further accepts a user operation to select the information representing the strength of the relationship between the multiple elements to be input into the generative model.

[0161] (Appendix 17) A program for causing a computer to execute the following processes: a process of acquiring information representing the strength of the relationship between multiple elements; a process of inputting the information representing the strength of the relationship between the multiple elements into a generative model; and a process of outputting information expressing the relationship between the multiple elements in language based on the analysis results by the generative model.

[0162] (Supplementary Note 18) The program according to Supplementary Note 17, wherein the information representing the strength of the relationship between the plurality of elements is numerical data representing the strength of the relationship between the plurality of elements.

[0163] (Supplementary Note 19) The program according to Supplementary Note 17, wherein the plurality of elements include a response variable that is a result and a plurality of explanatory variables that are causes.

[0164] (Supplementary Note 20) The program according to Supplementary Note 17, wherein the generative model is an information analysis program trained to contextualize input data and output output data according to the context of the input data.

[0165] (Supplementary Note 21) The program described in any one of Supplementary Notes 17 to 20, characterized in that the program causes the computer to execute a process of performing a causal analysis on information regarding the plurality of elements to obtain the information representing the strength of the relationship between the plurality of elements.

[0166] (Supplementary Note 22) The program described in any one of Supplementary Notes 17 to 20, characterized in that the computer is caused to execute a process of inputting the information representing the strength of the relationship between the multiple elements into the generative model in the form of a matrix whose elements are numbers representing the strength of the causal relationship between the multiple elements.

[0167] (Appendix 23) The program described in any one of Appendices 17 to 20, characterized in that the computer is caused to execute a process of outputting comments about the relationships between the multiple elements as the information based on the analysis results of the information representing the strength of the relationships between the multiple elements by the generative model.

[0168] (Supplementary Note 24) The program described in any one of Supplementary Notes 17 to 20, further comprising causing the computer to execute a process of accepting a user operation to select the information representing the strength of the relationship between the multiple elements to be input into the generative model.

[0169] (Supplementary Note 25) An information processing system comprising: an information processing device; one or more user terminals used by a user to input information or instructions to the information processing device; and an information analysis execution device trained to contextualize the input data and output output data according to the context of the input data, wherein the information processing device acquires information representing the strength of the relationship between a plurality of elements from the one or more user terminals, inputs the information representing the strength of the relationship between the plurality of elements to the information analysis execution device, and outputs information expressing the relationship between the plurality of elements in language based on the analysis results by the information analysis execution device.

[0170] (Supplementary Note 26) The information processing system according to Supplementary Note 25, wherein the information representing the strength of the relationship between the plurality of elements is numerical data representing the strength of the relationship between the plurality of elements.

[0171] (Supplementary Note 27) The information processing system according to Supplementary Note 25, wherein the plurality of elements includes a response variable that is a result and a plurality of explanatory variables that are causes.

[0172] (Supplementary Note 28) The information processing system according to Supplementary Note 25, wherein the generative model is an information analysis program trained to contextualize input data and output output data according to the context of the input data.

[0173] (Appendix 29) An information processing system described in any one of Appendices 25 to 28, characterized in that the computer obtains the information representing the strength of the relationship between the multiple elements by performing causal analysis on information regarding the multiple elements.

[0174] (Supplementary Note 30) The information processing system described in any one of Supplementary Notes 25 to 28, characterized in that the computer inputs the information representing the strength of the relationship between the multiple elements into the generative model in the form of a matrix whose elements are numbers representing the strength of the causal relationship between the multiple elements.

[0175] (Appendix 31) The information processing system described in any one of Appendices 25 to 28, characterized in that the information processing device outputs comments about the relationship between the multiple elements as the information based on the analysis results of the information representing the strength of the relationship between the multiple elements by the generative model.

[0176] (Appendix 32) The information processing system described in any one of Appendices 25 to 28, characterized in that the information processing device further accepts a user operation to select the information representing the strength of the relationship between the multiple elements to be input into the generative model.

[0177] The present disclosure has been described above with reference to several embodiments. However, the present disclosure is not limited to the above embodiments. Each embodiment can be combined with other embodiments as appropriate. Furthermore, various modifications that would be understood by a person skilled in the art can be made to the configurations and details of the above embodiments within the scope of the present disclosure. This application claims priority based on Japanese Patent Application No. 2024-006391, filed January 18, 2024, the entire disclosure of which is incorporated herein by reference.

[0178] The present disclosure can be used, for example, in information processing using generative models.

[0179] REFERENCE SIGNS LIST 1 Information processing system 10 Information processing device 11 Acquisition unit 12 Input unit 13 Output unit 20 Information processing device 24 Reception unit 30 Information processing device 31 Acquisition unit 32 Input unit 33 Output unit 100 User terminal 200 Information analysis execution device

Claims

1. An information processing apparatus comprising: an acquisition means for acquiring information representing the strength of relationships between a plurality of elements; an input means for inputting the information representing the strength of relationships between the plurality of elements into a generation model; and an output means for outputting information expressing in language the relationships between the plurality of elements based on the analysis result by the generation model.

2. The information processing apparatus according to claim 1, characterized in that the information representing the strength of relationships between the plurality of elements is numerical data or chart data representing the strength of relationships between the plurality of elements.

3. The information processing apparatus according to claim 1, characterized in that the plurality of elements include an objective variable that is a result and a plurality of explanatory variables that are causes.

4. The information processing apparatus according to claim 1, characterized in that the generation model is an information analysis program trained to contextualize input data and output output data according to the context of the input data.

5. The information processing apparatus according to any one of claims 1 to 4, characterized in that the acquisition means acquires the information representing the strength of relationships between the plurality of elements obtained by performing causal analysis on the information regarding the plurality of elements.

6. The information processing apparatus according to any one of claims 1 to 4, characterized in that the input means inputs the information representing the strength of relationships between the plurality of elements into the generation model in a matrix format having numerical values representing the strength of causal relationships between the plurality of elements as elements.

7. The information processing apparatus according to any one of claims 1 to 4, characterized in that the output means outputs a comment on the relationships between the plurality of elements as the information based on the analysis result of the information representing the strength of relationships between the plurality of elements by the generation model.

8. The information processing apparatus according to any one of claims 1 to 4, further comprising a reception means for receiving a user operation for selecting the information representing the strength of relationships between the plurality of elements to be input into the generation model.

9. An information processing method, wherein a computer acquires information representing the strength of relationships between a plurality of elements, inputs the information representing the strength of relationships between the plurality of elements into a generation model, and outputs information expressing in language the relationships between the plurality of elements based on the analysis result by the generation model.

10. The information processing method according to claim 9, wherein the information representing the strength of the relationship between the plurality of elements is numerical data representing the strength of the relationship between the plurality of elements.

11. The information processing method according to claim 9, wherein the plurality of elements include an objective variable that is a result and a plurality of explanatory variables that are causes.

12. The information processing method according to claim 9, wherein the generation model is an information analysis program trained to contextualize input data and output output data according to the context of the input data.

13. The information processing method according to any one of claims 9 to 12, wherein the computer obtains the information representing the strength of the relationship between the plurality of elements by performing causal analysis on the information regarding the plurality of elements.

14. The information processing method according to any one of claims 9 to 12, wherein the computer inputs the information representing the strength of the relationship between the plurality of elements into the generation model in the form of a matrix having numerical values representing the strength of the causal relationship between the plurality of elements as elements.

15. The information processing method according to any one of claims 9 to 12, wherein the computer outputs a comment on the relationship between the plurality of elements as the information based on the analysis result of the information representing the strength of the relationship between the plurality of elements by the generation model.

16. The information processing method according to any one of claims 9 to 12, wherein the computer further receives a user operation for selecting the information representing the strength of the relationship between the plurality of elements to be input into the generation model.

17. A non-transitory recording medium storing a program for causing a computer to execute a process of obtaining information representing the strength of the relationship between a plurality of elements, a process of inputting the information representing the strength of the relationship between the plurality of elements into a generation model, and a process of outputting information expressing the relationship between the plurality of elements in language based on the analysis result by the generation model.

18. The recording medium according to claim 17, wherein the information representing the strength of the relationship between the plurality of elements is numerical data representing the strength of the relationship between the plurality of elements.

19. The recording medium according to claim 17, wherein the plurality of elements include a target variable as a result and a plurality of explanatory variables as causes.

20. An information processing system comprising: an information processing device; one or more user terminals used by a user to input information or instructions to the information processing device; and an information analysis execution device trained to contextualize input data and output output data according to the context of the input data, wherein the information processing device obtains information representing the strength of the relationship between a plurality of elements from the one or more user terminals, inputs the information representing the strength of the relationship between the plurality of elements to the information analysis execution device, and outputs information expressing the relationship between the plurality of elements in language based on the analysis result by the information analysis execution device.

Images