Information processing device, information processing method, and program

Abstract

An information processing device 100 of the present disclosure comprises: a generation unit 121 for generating a graph that represents a plurality of types of preset text elements generated from sentence data, such representation being in accordance with the type of element and using nodes and edges connecting the nodes, and for generating a connection graph in which a plurality of graphs are connected in accordance with the content of nodes; and an extraction unit 122 for extracting a graph having a preset relationship among nodes on the basis of the connection graph.

Description

Information processing device, information processing method, and program

[0001] The present disclosure relates to an information processing device, an information processing method, and a program.

[0002] In order to utilize a large amount of text data, text data is structured. For example, Patent Literature 1 describes extracting subjects, predicates, and objects from text data and generating graph information that shows these in a graph structure.

[0003] International Publication No. 2021 / 205639

[0004] However, the technology described in Patent Document 1 only displays an overall image of multiple pieces of text data in a graph structure, which makes it difficult and time-consuming to analyze a huge amount of text data.

[0005] Therefore, an object of the present disclosure is to provide an information processing device that can solve the above-mentioned problem that analyzing a huge amount of text data is difficult and time-consuming.

[0006] An information processing device according to one embodiment of the present disclosure includes a generation unit that generates a graph in which a plurality of types of pre-set sentence elements generated from text data are represented by nodes and edges connecting the nodes according to the type of the elements, and generates a connected graph that connects the plurality of graphs according to the contents of the nodes; and an extraction unit that extracts the graphs having pre-set relationships between the nodes based on the connected graph.

[0007] Furthermore, an information processing method according to one embodiment of the present disclosure is configured to generate a graph in which a plurality of types of pre-set sentence elements generated from text data are represented by nodes and edges connecting the nodes according to the type of the elements, generate a connected graph in which a plurality of the graphs are connected according to the contents of the nodes, and extract, based on the connected graph, the graphs having pre-set relationships between the nodes.

[0008] Furthermore, a program according to one embodiment of the present disclosure has a configuration that causes a computer to execute the following processes: generate a graph representing a plurality of types of pre-set sentence elements generated from text data using nodes and edges connecting the nodes according to the type of the elements; generate a connected graph connecting the plurality of graphs according to the contents of the nodes; and extract, based on the connected graph, the graph having a pre-set relationship between the nodes.

[0009] With the above-described configuration, the present disclosure makes it possible to easily analyze a huge amount of text data in a short amount of time.

[0010] FIG. 1 is a block diagram showing the configuration of an information processing device according to a first embodiment of the present disclosure. FIG. 2 is a diagram showing the state of processing by the information processing device disclosed in FIG. 1. FIG. 3 is a diagram showing the state of processing by the information processing device disclosed in FIG. 1. FIG. 4 is a diagram showing the state of processing by the information processing device disclosed in FIG. 1. FIG. 5 is a diagram showing the state of processing by the information processing device disclosed in FIG. 1. FIG. 6 is a flowchart showing the operation of the information processing device disclosed in FIG. 1. FIG. 7 is a block diagram showing the hardware configuration of an information processing device according to a second embodiment of the present disclosure. FIG. 8 is a block diagram showing the configuration of an information processing device according to a second embodiment of the present disclosure.

[0011] First Embodiment A first embodiment of the present disclosure will be described with reference to Fig. 1 to Fig. 7. Fig. 1 is a diagram for explaining the configuration of an information processing device, and Fig. 2 to Fig. 7 are diagrams for explaining the processing operation of the information processing device.

[0012] [Configuration] The information processing device 10 in this embodiment displays text data in a graph structure. In particular, the information processing device 10 targets text data contained in the text body or attachments of emails, messages, SNS (Social Networking Service) posts, etc., and displays and analyzes this text data in a graph structure. As an example, in this embodiment, emails that may be related to incidents that are the subject of police investigation are targeted and displayed in a graph structure and analyzed. Note that the text data that is displayed in a graph structure and analyzed is not limited to the above-mentioned emails, etc., and may be any text data.

[0013] The information processing device 10 is composed of one or more information processing devices each including a calculation device and a storage device. As shown in FIG. 1 , the information processing device 10 includes an input unit 11, a preprocessing unit 12, a graph generation unit 13, a graph analysis unit 14, and a display control unit 15. The functions of the input unit 11, the preprocessing unit 12, the graph generation unit 13, the graph analysis unit 14, and the display control unit 15 can be realized by the calculation device executing a program for realizing each function stored in the storage device. The information processing device 10 also includes a text data storage unit 16 and a graph storage unit 17. The text data storage unit 16 and the graph storage unit 17 are each composed of a storage device. A display device 30 such as a display is also connected to the information processing device 10. Each component will be described in detail below.

[0014] The input unit 11 accepts input of text data to be processed and stores it in the text data storage unit 16. For example, the input unit 11 acquires emails stored in another information processing device or storage medium as text data and stores each item in the text data storage unit 16. At this time, it is assumed that the email contains data such as a "header" including a "subject," "sender (From)," "destination (To)," and "sent time (Time)," and a "body." However, the email may contain other information, and may also contain an "attachment" containing text data.

[0015] The preprocessing unit 12 reads emails one by one from the storage device and preprocesses the text data included in the emails. In this embodiment, the preprocessing unit 12 divides the text data, such as the "body" and "attachments" included in the email, into sentences and extracts and generates a subject, predicate, and object for each sentence. For example, the preprocessing unit 12 divides the text data into sentences based on preset criteria and extracts and generates a subject, predicate, and object for each sentence using other text data before and after the sentence and other information in the email (e.g., the subject, sender, and recipient included in the header). At this time, the preprocessing unit 12 extracts the subject, predicate, and object from each sentence using preset text analysis rules and analysis models. As an example, the preprocessing unit 12 can extract the subject, predicate, and object from a sentence using a technology called Open Information Extraction (OpenIE).

[0016] The extraction of sentence elements such as the subject, predicate, and object described above may be achieved using a machine learning model that extracts sentence elements from a sentence. For example, by inputting a target sentence into a machine learning model that extracts sentence elements from a pre-generated sentence, the sentence elements, that is, the subject, predicate, and object, are extracted and output. For example, such a machine learning model is generated by supervised learning. In this case, supervised learning can generate the above-mentioned machine learning model by dividing text data into sentences and using training data consisting of each sentence and a pair of the subject, predicate, and object of the sentence.

[0017] Furthermore, when a subject or object is omitted from a sentence, the preprocessing unit 12 uses other preceding and following sentence data or other information in the email (e.g., a header) to newly complete the subject and object, and extracts and generates a subject, predicate, and object from the sentence. In this case, the preprocessing unit 12, as described above, uses preset sentence analysis rules and analysis models or a technology called OpenIE to partially complete the subject, predicate, and object from the sentence, and extracts and generates the subject, predicate, and object. The preprocessing unit 12 then associates the set of subject, predicate, and object extracted from the sentence as described above with the email from which they were extracted.

[0018] The subject and object completion described above may be achieved using a machine learning model that generates sentences from sentences. For example, by inputting the sentence to be completed, the sentences before and after it, and header information containing sender and recipient information associated with emails or social media posts into a machine learning model that generates sentences from previously generated sentences, a sentence in which the omitted sentence elements, such as the subject and object, are completed for the sentence to be completed is output. For example, such a machine learning model is created using supervised learning. In this case, supervised learning divides text data into sentences and uses training data consisting of a sentence in which sentence elements such as the subject and object are omitted, the preceding and following sentences, header information, and the completed sentence. Note that instead of the preceding and following sentences, sentence elements extracted using a machine learning model that extracts sentence elements from the previous sentence may be used. Furthermore, completion of elements such as the subject and object may be performed not only by using a machine learning model that generates sentences from sentences, but also by adding sentence elements such as the subject and object extracted using a machine learning model that extracts sentence elements from the previous sentence to the beginning or end of the target sentence.

[0019] While the above example illustrates a case where text data is divided into sentences based on preset criteria, a sentence may be a character string of any length, and may be divided into sentences based on any criteria. Furthermore, while the above example illustrates a case where one of each of three types of sentence elements, i.e., subject, predicate, and object, is generated for each sentence of text data, any number of types of sentence elements may be generated, and multiple elements of each type may be generated. For example, the preprocessing unit 12 may generate elements such as subject, predicate, object, complement, and modifier.

[0020] The graph generation unit 13 (generation unit) generates a graph in which the subject, predicate, and object generated for each sentence of the text data as described above are represented in a graph structure consisting of nodes and edges connecting the nodes. Specifically, as shown in FIG. 2 (2-1), the graph generation unit 13 generates a graph g in which the "subject" and "object" are each represented by a node, and the "predicate" is represented by an edge connecting the nodes. Note that the graph is not limited to the structure described above, and the node and edge structures may vary depending on the number and types of elements in the sentence. In other words, the graph is not limited to one represented by two nodes and one edge connecting them as shown in FIG. 2 (2-1), and additional nodes and edges may be added.

[0021] Furthermore, the graph generation unit 13 connects multiple graphs according to the content of the nodes of each graph, generates a connected graph, and stores the connected graph in the graph storage unit 17. Specifically, the graph generation unit 13 connects multiple graphs at the locations of nodes with the same content included in each of the multiple graphs to generate a connected graph. Here, an example of connecting graphs to generate a connected graph by the graph generation unit 13 will be described with reference to FIG. 2 (2-2). First, as shown in the left diagram of FIG. 2 (2-2), it is assumed that two graphs g1 and g2 have been generated. In this case, the "object" nodes of the two graphs g1 and g2 have the same content, "object 1." Therefore, the graph generation unit 13 connects the two graphs g1 and g2 at the location of the "object 1" node so that the two graphs g1 and g2 share the "object 1" node. That is, as shown in the right diagram of FIG. 2 (2-2), the graph generation unit 13 generates a connected graph G by connecting the edge "Predicate 1" connected to the node "Subject 1" of graph g1 and the edge "Predicate 2" connected to the node "Subject 2" of graph g2 to one node "Object 1." Note that the connected graph G is not limited to connecting two graphs g1 and g2 as described above, and more graphs may be connected. Furthermore, the graph generation unit 13 is not limited to connecting graphs by the above-described method, and may generate a connected graph by connecting graphs by other methods. For example, if the contents of nodes in multiple graphs are not completely identical but can be determined to be identical or related based on predetermined criteria, the graph generation unit 13 may connect different graphs at such nodes. Furthermore, the graph generation unit 13 may connect multiple graphs at the same node in the same manner as described above in response to a user operation, for example, an operation to specify the same node for different graphs.

[0022] FIG. 3 shows an example of the connected graph G generated by the graph generation unit 13. FIG. 3 illustrates the connected graph G displayed on the display device 30 by the display control unit 15. As shown in FIG. 3, the display control unit 15 displays a display screen divided into multiple areas on the screen of the display device 30, for example, a graph display area D1, a mail list display area D2, a mail body display area D3, and a slide bar display area D4. The display control unit 15 then displays the generated connected graph G in the graph display area D1. The connected graph G in this example includes, for example, a graph g11 surrounded by a dotted line, which includes the subject node "Minami," the predicate edge "setting," and the object node "Store AA," and a graph g12 surrounded by a dotted line, which includes the subject node "Okumura," the predicate edge "I will visit you," and the object node "Store AA." These graphs are connected by the object node "Store AA."

[0023] The display control unit 15 displays the connected graph G in the graph display area D1 as described above, and also controls other display areas on the display screen to display information as follows: As shown in Fig. 3, the display control unit 15 displays in the email list display area D2 a list of header information of emails that contain a sentence that is the source of the graph included in the connected graph G, that is, a list of the "Subject" (title), "Sender (From)", "Destination (To)", and "Sent Time" of the email. Note that the display control unit 15 may also display a list of only the "Subject" of the email in the email list display area D2.

[0024] The display control unit 15 also displays email data in the email body display area D3, as shown in FIG. 3 . For example, the display control unit 15 displays the "Subject" (title), "From," "To," "Time," and "Body" of an email selected by the user, as shown in gray in FIG. 3 , among the emails whose "Subjects" are listed in the email list display area D2, in the email body display area D3. Note that the "Time" may be the time the email was sent or received. The display control unit 15 may also display text data included in an "attachment" of the email in the email body display area D3, or may display only the text data of the "Body" and "Attachment" without displaying the "Header" information.

[0025] Furthermore, the display control unit 15 displays a histogram showing the number of emails per hour in the slide bar display area D4, as shown in Fig. 3. Specifically, the display control unit 15 displays a histogram with time (day) on the horizontal axis and the number of emails sent on the vertical axis. In this case, a slide bar B is set for time on the horizontal axis, and the user can specify time by changing the position and length of the slide bar B, as shown by the thick line in Fig. 3. The display control unit 15 then displays a list of emails located within the range of the slide bar B in the email list display area D2.

[0026] The graph analysis unit 14 (extraction unit) extracts a graph having predetermined relationships between nodes from the connected graph G generated as described above. For example, the graph analysis unit 14 extracts a graph consisting of a specific node, an edge connecting the specific node, and another node connected to the specific node by an edge. In this case, the specific node is assumed to be a node searched for or specified by a user on the display screen of the display device 30. For example, the user can search for a node corresponding to the keyword by entering a keyword in a search field displayed on the display screen of the display device 30, or can specify a node by selecting it with a pointer on the connected graph G displayed in the graph display area D1. Then, the display control unit 15 described above controls the display of the graph extracted by the graph analysis unit 14 on the display screen of the display device 30 so that it is displayed separately from other graphs.

[0027] As an example, as shown in FIG. 4 , if the node “Store AA” is designated as a specific node in the connected graph G, the graph analysis unit 14 extracts a graph including all edges and other nodes connected to the specific node “Store AA.” The display control unit 15 then displays the extracted graph distinguishably from other graphs, as indicated by the bold lines in FIG. 4 . In the example of FIG. 4 , a graph including the node “Store AA,” the edge “Setting,” and the node “Minami,” a graph including the node “Store AA,” the edge “I’ll Visit You,” and the node “Okumura,” and a graph including the node “Store AA,” the edge “Like,” and the node “Kawashima” are displayed with bold lines to distinguish them from other graphs. Note that in the example of FIG. 4 , the display control unit 15 displays the extracted graphs in gray to make them less noticeable, thereby emphasizing them relative to the other graphs.

[0028] 4, people, places, etc. correspond to the nodes that serve as subjects and objects. Therefore, by highlighting the extracted graph on the connected graph G as described above, it is possible to easily recognize the relationships between people and the relationships between people and places. For example, when analyzing emails in a police investigation, by specifying a node that corresponds to a person or place of interest, it is possible to easily recognize other people, places, etc. that may be related to the person or place of that node.

[0029] Another example is shown in FIG. 5 . In this example, assume that the node of "tampering," a term that may be related to a crime, is specified in the connected graph G as a node of interest in the investigation. The graph analysis unit 14 then extracts a graph that includes all edges and other nodes connected to the specific node "tampering." The display control unit 15 then displays the extracted graph, as indicated by the bold lines in FIG. 5 , distinguishing it from other graphs. That is, in the example of FIG. 5 , a graph including the node "Minami" connected to the node "tampering" by an edge, a graph including the node "Okumura" connected to the node "tampering" by an edge, a graph including the node "Sawada" connected to the node "tampering" by an edge, and so on, are displayed with bold lines to distinguish them from other graphs.

[0030] Furthermore, the display control unit 15 displays the email from which the extracted graph was generated, distinguishing it from other emails. Specifically, as shown in FIGS. 4 and 5 , the display control unit 15 displays a black circle in the subject line of the email from which the extracted graph was generated in the email list display area D2. This allows the user to easily identify the email corresponding to the extracted graph. When the user selects the subject line of an email in the email list display area D2, the display control unit 15 displays the body of the selected email in the email body display area D3. At this time, the display control unit 15 displays text data, which is a sentence from the email from which the extracted graph was generated, distinguished from other text data in the email body display area D3. For example, in the example of FIG. 4 , the display control unit 15 indicates that the email selected in the email list display area D2 is the email from which the extracted graph was generated, and displays the sentence "I will set up store AA," from which the extracted graph was generated, underlined.

[0031] Furthermore, as shown in FIG. 3 , the display control unit 15 displays a histogram representing the number of emails per time period in the slide bar display area D4. Specifically, the display control unit 15 displays a histogram with the time (day) on the horizontal axis and the number of emails sent on the vertical axis. A slide bar B is set for the time on the horizontal axis, and the user can specify the time by changing the length of the slide bar B, as indicated by the bold line in FIG. 3 . The display control unit 15 then displays a list of emails located within the range of the slide bar B in the email list display area D2. Specifically, as shown in FIG. 6 , when the length of the slide bar B is set by a user operation, the display control unit 15 displays the "subject" of an email corresponding to a time within the range of the slide bar B in bold, based on the time associated with the email, to distinguish it from the "subject" of other emails. Note that in the example of FIG. 6 , the display control unit 15 displays the "subject" of emails other than those located within the range of the slide bar B in gray to make them less noticeable, thereby emphasizing the "subject" of the email located within the range of the slide bar B relative to the other emails. At this time, if there is an extracted graph as described above, the display control unit 15 displays the "subject" of the email corresponding to the graph generated from the email located within the range of slide bar B among the extracted graphs in bold, as shown in Figure 7, so as to distinguish it from other subjects.

[0032] The display control unit 15 also displays graphs generated from emails located within the range of the slide bar B in the graph display area D1. Specifically, as shown in FIG. 6 , when the length of the slide bar B is set by a user operation, the display control unit 15 displays graphs generated from emails corresponding to times within the range of the slide bar B in bold, based on the times associated with the emails, to distinguish them from graphs generated from other emails. In the example of FIG. 6 , the display control unit 15 displays graphs other than those generated from emails located within the range of the slide bar B in gray to make them less noticeable, thereby highlighting the graphs generated from emails located within the range of the slide bar B relative to the other graphs. At this time, if there are any extracted graphs as described above, the display control unit 15 displays only the graphs generated from emails located within the range of the slide bar B in bold among the extracted graphs to distinguish them from the other graphs, as shown in FIG. 7 .

[0033] Then, as described above, each time the length or position of the slide bar B is changed and specified by a user operation, the display control unit 15 changes the list of emails to be highlighted in the email list display area D2 or changes the graph to be highlighted in the graph display area D1 according to the specified time, allowing the user to easily recognize the chronological changes in the exchange of emails between each person.

[0034] [Operation] Next, the operation of the information processing device 10 described above will be described with reference to the flowchart of FIG.

[0035] The information processing device 10 receives input of text data to be processed and stores it in the text data storage unit 16 (step S1). In this embodiment, the information processing device receives input of emails that may be related to a case that is the subject of a police investigation as text data. At this time, the email is assumed to include data such as a "header" including a "subject," a "sender (From)," a "destination (To)," and a "time of sending," as well as a "body."

[0036] Next, the information processing device 10 reads each email from the storage device and performs preprocessing on the text data contained in the email. In this embodiment, as preprocessing, the information processing device 10 divides the text data of the "body" contained in the email into sentences and extracts the subject, predicate, and object for each sentence. Prior to this, if the subject and object are omitted in a sentence, the information processing device 10 uses other text data before and after or other information in the email (e.g., the header) to newly complete the subject and object (step S2). Then, the information processing device 10 extracts the subject, predicate, and object for each sentence in which the subject and object have been completed and for each sentence in which completion is not required (step S3). The information processing device 10 then associates the extracted sets of subject, predicate, and object with the email from which they were extracted (step S4).

[0037] Next, the information processing device 10 generates a graph in which the subjects, predicates, and objects generated for each sentence of the text data are represented in a graph structure consisting of nodes and edges connecting the nodes. Furthermore, the information processing device 10 connects multiple graphs according to the contents of the nodes in each graph to generate a connected graph. The information processing device 10 then displays the connected graph G on the display screen of the display device 30 (step S5). For example, as shown in FIG. 3 , the information processing device 10 displays the generated connected graph G in a graph display area D1 on the display screen of the display device 30.

[0038] At this time, the information processing device 10 displays a list of header information of emails that contain the sentence that is the source of the graph included in the connected graph G in an email list display area D2 on the display screen. The information processing device 10 also displays the body of the selected email in an email body display area D3 on the display screen. Furthermore, the information processing device 10 displays a histogram showing the number of emails by time in a slide bar display area D4 on the display screen.

[0039] The information processing device 10 then accepts a user's operation on the displayed connected graph and analyzes the graph. For example, when the information processing device 10 accepts a node designation by a user's operation, it extracts a graph related to the designated node. The information processing device 10 then highlights the extracted graph to distinguish it from other graphs (step S6). For example, as shown in FIG. 4 , if the user designates the node "Store AA" in the connected graph G as a specific node, the information processing device 10 extracts a graph including all edges and other nodes connected to the specific node "Store AA," and highlights the extracted graph to distinguish it from other graphs, as indicated by the bold line in FIG. 4 .

[0040] At this time, the information processing device 10 displays the email from which the extracted graph was generated as described above, distinguished from other emails. For example, as shown in FIGS. 4 and 5 , the information processing device 10 marks the "subject" of the email from which the extracted graph was generated with a black circle in the email list display area D2. Furthermore, when the user selects the "subject" of an email in the email list display area D2, the information processing device 10 displays the "body" of the selected email in the email body display area D3. For example, the information processing device 10 displays text data, which is a sentence from the email from which the extracted graph was generated, distinguished from other text data in the email body display area D3.

[0041] In this way, the information processing device 10 of this embodiment highlights the extracted graph, allowing the user to easily recognize the relationships between people and the relationships between people and places. Furthermore, in this embodiment, the user can easily recognize the email corresponding to the extracted graph, and can also easily recognize the text of the email.

[0042] Furthermore, the information processing device 10 receives a user operation on the slide bar B displayed in the slide bar display area D4 and controls the display of the graph and email list. For example, as shown by the thick line in Figures 6 and 7, when the length of the slide bar B is changed by a user operation, the information processing device 10 highlights the list of emails located within the range of the slide bar B, or highlights a graph generated from the emails located within the range of the slide bar B.

[0043] In this way, in the information processing device 10 of this embodiment, the list of emails and graphs to be highlighted are changed each time the length or position of the slide bar B is changed and specified, thereby enabling the user to easily recognize the chronological changes in the exchange of emails between people.

[0044] <Embodiment 2> Next, a second embodiment of the present disclosure will be described with reference to Fig. 9 and Fig. 10. Fig. 9 and Fig. 10 are block diagrams showing the configuration of an information processing device in embodiment 2. Note that this embodiment shows an outline of the configuration of the information processing device described in the above-mentioned embodiment.

[0045] First, the hardware configuration of the information processing device 100 in this embodiment will be described with reference to Fig. 9. The information processing device 100 is configured as a general information processing device, and is equipped with the following hardware configuration, for example: CPU (Central Processing Unit) 101 (arithmetic unit); ROM (Read Only Memory) 102 (storage device); RAM (Random Access Memory) 103 (storage device); programs 104 loaded into RAM 103; a storage device 105 that stores the programs 104; a drive device 106 that reads and writes data from and to a storage medium 110 external to the information processing device; a communication interface 107 that connects to a communication network 111 external to the information processing device; an input / output interface 108 that inputs and outputs data; and a bus 109 that connects the various components.

[0046] 9 shows an example of the hardware configuration of the information processing device 100, and the hardware configuration of the information processing device is not limited to the above-described case. For example, the information processing device may be configured with only a part of the above-described configuration, such as excluding the drive device 106. Furthermore, the information processing device may use a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), an MPU (Micro Processing Unit), an FPU (Floating Point Number Processing Unit), a PPU (Physics Processing Unit), a TPU (Tensor Processing Unit), a quantum processor, a microcontroller, or a combination thereof, instead of the above-described CPU.

[0047] The information processing device 100 can then configure and include the generation unit 121 and extraction unit 122 shown in FIG. 10 by having the CPU 101 acquire and execute the program group 104. The program group 104 is stored in advance in the storage device 105 or the ROM 102, for example, and is loaded into the RAM 103 and executed by the CPU 101 as needed. The program group 104 may also be supplied to the CPU 101 via the communication network 111, or may be stored in advance in the storage medium 110, with the drive device 106 reading out the programs and supplying them to the CPU 101. However, the generation unit 121 and extraction unit 122 described above may also be configured using dedicated electronic circuits for realizing such means.

[0048] The generation unit 121 generates a graph in which a plurality of types of sentence elements set in advance, generated from text data, are represented by nodes according to the types of the elements and edges connecting the nodes, and also generates a connected graph in which the plurality of graphs are connected according to the contents of the nodes. For example, the generation unit 121 generates a connected graph by connecting the plurality of graphs at the locations of nodes with the same contents included in the plurality of graphs.

[0049] The extraction unit 122 extracts a graph having predetermined relationships between nodes based on the connection graph. For example, the extraction unit 122 extracts a graph consisting of specific nodes designated by the user, edges connecting to the specific nodes, and other nodes connected to the specific nodes by edges.

[0050] With the above-described configuration, the present disclosure extracts a graph in which nodes have relationships from a connected graph generated from a sentence. This makes it possible to easily recognize relationships between people included in the extracted graph, relationships between people and places, etc. As a result, even a huge amount of sentence data can be easily analyzed in a short time.

[0051] The above-described program can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / Ws, and semiconductor memories (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). The program may also be supplied to a computer by various types of transitory computer-readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. The transitory computer-readable media can be supplied to a computer via wired communication paths such as electric wires and optical fibers, or via wireless communication paths.

[0052] Although the present disclosure has been described above with reference to the above-described embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, at least one or more of the functions of the generation unit 121 and the extraction unit 122 described above may be executed by an information processing device installed and connected anywhere on a network, that is, may be executed by so-called cloud computing.

[0053] <Supplementary Notes> Some or all of the above embodiments may be described as in the following supplementary notes. Below, an outline of the configurations of an information processing method, an information processing device, and a program according to the present disclosure will be described. However, the present disclosure is not limited to the following configurations. (Supplementary Note 1) An information processing device comprising: a generation unit that generates graphs in which a plurality of types of preset sentence elements generated from text data are represented by nodes and edges connecting the nodes according to the types of the elements, and that generates a connected graph connecting the plurality of graphs according to the contents of the nodes; and an extraction unit that extracts the graphs having a predetermined relationship between the nodes based on the connected graph. (Supplementary Note 2) The information processing device according to Supplementary Note 1, wherein the generation unit generates the connected graph by connecting the plurality of graphs at locations of the nodes having the same content included in each of the plurality of graphs, and the extraction unit extracts the graph consisting of a specific node, the edge connecting the specific node, and another node connected to the specific node by the edge. (Supplementary Note 3) The information processing device according to Supplementary Note 2, comprising a display control unit that displays the connected graph on a display device and displays the extracted graph on the display device distinguished from other graphs. (Supplementary Note 4) The information processing device according to Supplementary Note 3, wherein the display control unit displays the extracted graph distinguished from other graphs based on time information associated with the text data from which the extracted graph was generated. (Supplementary Note 5) The information processing device according to Supplementary Note 3, wherein the display control unit displays the graph generated from the text data associated with time information corresponding to a specified time distinguished from the other graphs. (Supplementary Note 6) The information processing device according to Supplementary Note 3, wherein the display control unit displays the text data from which the extracted graph was generated on the display device.(Supplementary Note 7) The information processing device according to Supplementary Note 6, wherein the display control unit displays a list of titles associated with the text data on the display device, and displays the title associated with the text data from which the extracted graph was generated distinguished from the other titles. (Supplementary Note 8) The information processing device according to Supplementary Note 7, wherein the display control unit displays the title associated with the text data from which the extracted graph was generated distinguished from the other titles based on time information associated with the text data from which the extracted graph was generated. (Supplementary Note 9) The information processing device according to Supplementary Note 7, wherein the display control unit displays the title associated with the text data associated with time information corresponding to a specified time distinguished from the other titles. (Supplementary Note 10) An information processing method comprising: generating a graph in which a plurality of preset types of sentence elements generated from text data are represented by nodes and edges connecting the nodes according to the type of the elements; generating a connected graph by connecting the plurality of graphs according to the content of the nodes; and extracting the graph having a preset relationship between the nodes based on the connected graph. (Supplementary Note 11) An information processing method according to Supplementary Note 10, comprising: generating the connected graph by connecting a plurality of graphs at the locations of the nodes with the same content included in each of the graphs; and extracting the graph consisting of a specific node, the edge connecting to the specific node, and another node connected to the specific node by the edge. (Supplementary Note 12) An information processing method according to Supplementary Note 11, comprising: displaying the connected graph on a display device; and displaying the extracted graph on the display device distinguished from other graphs. (Supplementary Note 13) The information processing method according to Supplementary Note 12, wherein the extracted graph is displayed in a manner distinguishable from other graphs based on time information associated with the text data from which the extracted graph was generated.(Supplementary Note 14) The information processing method according to Supplementary Note 12, wherein the graph generated from the text data associated with time information corresponding to a specified time is displayed distinguished from the other graphs. (Supplementary Note 15) The information processing method according to Supplementary Note 12, wherein the text data from which the extracted graph was generated is displayed on the display device. (Supplementary Note 16) The information processing method according to Supplementary Note 15, wherein a list of titles associated with the text data is displayed on the display device, and the title associated with the text data from which the extracted graph was generated is displayed distinguished from the other titles. (Supplementary Note 17) The information processing method according to Supplementary Note 16, wherein the title associated with the text data from which the extracted graph was generated is displayed distinguished from the other titles, based on time information associated with the text data from which the extracted graph was generated. (Supplementary Note 18) The information processing method according to Supplementary Note 17, wherein the title associated with the text data associated with time information corresponding to a specified time is displayed in a manner distinguished from the other titles. (Supplementary Note 19) A computer-readable storage medium storing a program that causes a computer to execute the steps of: generating a graph in which a plurality of types of preset sentence elements generated from text data are represented by nodes and edges connecting the nodes according to the types of the elements, generating a connected graph in which a plurality of the graphs are connected according to the contents of the nodes, and extracting the graphs having a predetermined relationship between the nodes based on the connected graph.

[0054] REFERENCE SIGNS LIST 10 Information processing device 11 Input unit 12 Preprocessing unit 13 Graph generation unit 14 Graph analysis unit 15 Display control unit 16 Text data storage unit 17 Graph storage unit 100 Information processing device 101 CPU 102 ROM 103 RAM 104 Program group 105 Storage device 106 Drive device 107 Communication interface 108 Input / output interface 109 Bus 110 Storage medium 111 Communication network 121 Generation unit 122 Extraction unit

Claims

1. a generation unit that generates a graph in which a plurality of types of sentence elements set in advance, generated from the text data, are represented by nodes and edges connecting the nodes according to the types of the elements, and that generates a connected graph in which a plurality of the graphs are connected according to the contents of the nodes; an extraction unit that extracts the graph having a predetermined relationship between the nodes based on the connection graph; An information processing device comprising:

2. 2. The information processing device according to claim 1, the generation unit generates the connected graph by connecting the plurality of graphs at the locations of the nodes having the same content that are included in each of the plurality of graphs; the extraction unit extracts the graph consisting of a specific node, the edges connected to the specific node, and other nodes connected to the specific node by the edges; Information processing device.

3. 3. The information processing device according to claim 2, a display control unit that displays the connection graph on a display device and displays the extracted graph on the display device in a manner that distinguishes it from other graphs; Information processing device.

4. 4. The information processing device according to claim 3, the display control unit displays the extracted graph in a manner distinguished from other graphs based on time information associated with the text data from which the extracted graph was generated. Information processing device.

5. 4. The information processing device according to claim 3, the display control unit displays the graph generated from the text data associated with time information corresponding to a specified time in a manner distinguishable from other graphs. Information processing device.

6. 4. The information processing device according to claim 3, the display control unit displays the extracted text data, which is the source of generating the graph, on the display device. Information processing device.

7. 7. The information processing device according to claim 6, the display control unit displays a list of titles associated with the text data on the display device, and displays the title associated with the text data from which the extracted graph was generated, distinguished from other titles. Information processing device.

8. 8. The information processing device according to claim 7, the display control unit displays the title associated with the original text data from which the extracted graph was generated, distinguished from other titles, based on time information associated with the original text data from which the extracted graph was generated. Information processing device.

9. generating a graph in which a plurality of types of sentence elements set in advance, generated from the text data, are represented by nodes and edges connecting the nodes according to the types of the elements, and generating a connected graph in which a plurality of the graphs are connected according to the contents of the nodes; extracting the graph having a predetermined relationship between the nodes based on the connection graph; Information processing methods.

10. generating a graph in which a plurality of types of sentence elements set in advance, generated from the text data, are represented by nodes and edges connecting the nodes according to the types of the elements, and generating a connected graph in which a plurality of the graphs are connected according to the contents of the nodes; extracting the graph having a predetermined relationship between the nodes based on the connection graph; A program that causes a computer to perform a process.