Analytical device, analytical method, and program

Abstract

To appropriately understand an interrelationship of documents and characteristics of a document group.SOLUTION: An analyzing device comprises: a link connection unit that connects between documents by means of links and generates an edge list indicating a list of links for a document unit on the basis of a relationship of citing / being cited of a document group including a plurality of documents; a community analysis unit that divides the plurality of documents into communities being document groups of a lower hierarchy of the document group to make a portion being coarse for coarseness and denseness of the link as a boundary by performing network analysis for the edge list of the document unit generated by the link connection unit, and identifies an edge list of the community unit from the links of the document unit connecting between the communities; and an output control unit that outputs output information showing an illustration of a relationship between the communities on the basis of the communities and the edge list of the community unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an analysis device, an analysis method, and a program. [Background technology]

[0002] In recent years, a technique has become known for analyzing the citing and cited relationships of documents such as patent documents as a network diagram, thereby visualizing the interrelationships between documents in a set of documents (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Special Publication No. 2015-532495 Summary of the Invention [Problem to be solved by the invention]

[0004] However, with the above-mentioned conventional techniques, as the number of documents increases, the network diagram becomes complex, and it may become difficult to properly grasp the interrelationships between documents and the characteristics of the document group.

[0005] The present disclosure has been made to solve the above problem, and its purpose is to provide an analysis device, an analysis method, and a program that can appropriately grasp the interrelationships between documents and the characteristics of a group of documents. [Means for solving the problem]

[0006] In order to solve the above problem, one aspect of the present disclosure includes a link combining unit that combines documents with links based on citing / cited relationships of a document group including a plurality of documents and generates an edge list showing a list of links for each document; a community analysis unit that performs network analysis on the edge list for each document generated by the link combining unit to divide the plurality of documents into communities, which are document groups at a lower level of the document group, so that portions with sparse links serve as boundaries, and identifies edge lists for each community from the document links connecting the communities; and an output control unit that outputs output information illustrating relationships between the communities based on the communities and the edge lists for each community. a heat map generating unit that generates a heat map diagram for each community based on the number of documents included in the community of the document group and an edge list for each community; and a network generating unit that generates a network diagram for each community illustrated by nodes indicating the community of the document group and directional links based on the edge list for each community. Equipped with The output control unit outputs the output information in which the network diagram generated by the network generation unit is superimposed on the heat map diagram generated by the heat map generation unit for each community. It is an analytical device.

[0007] Furthermore, one aspect of the present disclosure includes a link combining step in which a link combining unit combines documents with links based on citing / cited relationships in a document group including a plurality of documents, and generates an edge list showing a list of links for each document; a community analysis step in which a community analysis unit performs network analysis on the edge list for each document generated by the link combining step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that portions with sparse links serve as boundaries, and identifies edge lists for each community from the document links connecting the communities; and an output control step in which an output control unit outputs output information illustrating relationships between the communities based on the communities and the edge lists for each community. a heat map generation step in which a heat map generation unit generates a heat map diagram for each community based on the number of documents included in the community of the document group and an edge list for each community; and a network generation step in which a network generation unit generates a network diagram for each community illustrated by nodes indicating the community of the document group and directional links based on the edge list for each community. Including and in the output control step, the output control unit outputs the output information in which the network diagram for each community generated in the network generation step is superimposed on the heat map diagram for each community generated in the heat map generation step. It is an analytical method.

[0008] Furthermore, one aspect of the present disclosure provides a method for processing a plurality of documents in a computer, the method comprising: a linking step of linking documents with links based on citing / cited relationships of the documents in a document group including the documents, and generating an edge list showing a list of links for each document; a community analysis step of performing network analysis on the edge list for each document generated by the linking step, thereby dividing the documents into communities that are document groups at a lower level of the document group so that portions with sparse links serve as boundaries, and identifying edge lists for each community from the document links connecting the communities; and an output control step of outputting output information illustrating relationships between the communities based on the communities and the edge lists for each community. a heat map generation step in which a heat map generation unit generates a heat map diagram for each community based on the number of documents included in the community of the document group and an edge list for each community; and a network generation step in which a network generation unit generates a network diagram for each community illustrated by nodes indicating the community of the document group and directional links based on the edge list for each community. A program to run and in the output control step, the output control unit outputs the output information in which the network diagram for each community generated in the network generation step is superimposed on the heat map diagram for each community generated in the heat map generation step. is. [Effects of the Invention]

[0009] According to the present disclosure, it is possible to appropriately grasp the interrelationships between documents and the characteristics of a group of documents. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a block diagram illustrating an example of an analysis device according to a first embodiment. [Figure 2] 1A and 1B are diagrams illustrating a network diagram for patent documents and an example of community division in the first embodiment. [Figure 3] FIG. 2 is a diagram illustrating an example of a network diagram in units of communities according to the first embodiment. [Figure 4] FIG. 2 is a diagram showing an example of a network diagram in community units using a minimum spanning tree in the first embodiment. [Figure 5] 10 is a flowchart illustrating an example of a process for generating a community of analysis devices according to the first embodiment. [Figure 6] 10 is a flowchart showing an example of a process for generating a network diagram for each community by the analysis device according to the first embodiment. [Figure 7] FIG. 2 is a diagram showing an example of a network diagram with keywords for each community using a minimum spanning tree in the first embodiment. [Figure 8] For comparison, FIG. 10 is a diagram showing an example of a network diagram for each patent document. [Figure 9] FIG. 2 is a diagram illustrating an example of a network diagram in units of communities according to the first embodiment. [Figure 10] FIG. 10 is a block diagram illustrating an example of an analysis device according to a second embodiment. [Figure 11] FIG. 11 is a diagram illustrating an example of a heat map diagram in units of communities according to the second embodiment. [Figure 12] FIG. 11 is a diagram illustrating an example in which a network diagram in units of a community is superimposed on a heat map diagram in units of a community according to the second embodiment. [Figure 13] 10 is a flowchart showing an example of a process for generating a heat map diagram for each community in the analysis device according to the second embodiment. [Figure 14] For comparison, FIG. 10 is a diagram showing an example of a heat map diagram for patent documents. [Figure 15] FIG. 11 is a diagram illustrating an example of a heat map diagram in units of communities according to the second embodiment. [Figure 16] FIG. 10 is a diagram showing a first comparative example of a heat map diagram in units of communities in the second embodiment. [Figure 17] FIG. 10 is a diagram showing a second comparative example of a heat map diagram in units of communities in the second embodiment. [Figure 18] FIG. 10 is a diagram showing a third comparative example of a heat map diagram in units of communities in the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] An analysis device, an analysis method, and a program according to an embodiment of the present disclosure will be described below with reference to the drawings.

[0012] Before describing individual embodiments, the characteristics of the target document group will be described. The documents belonging to the target document group are linked to each other through relationships such as citation and cited relationships. In this disclosure, this linking is called a link. Since related documents are connected by links, a collection of related documents is extracted by following all the links. In this disclosure, this collection is called a cluster. The target document group will contain one or more clusters.

[0013] This cluster can be further divided into sub-clusters (communities) with stronger connections, taking into account the density of links. In other words, if a relatively dense link area is extracted as a community, the density of links between documents within the community will be higher than the density of links between documents within the community and documents outside the community, and it can be considered that a group of documents that can be distinguished from the outside has been extracted. These extraction methods are common and are called cluster extraction (cluster analysis) and community extraction (community analysis), respectively. In this disclosure, a hierarchical structure is considered in which a collection (community) is formed by a group of documents, and a group of these communities is formed by a group of even larger collections (groups of target documents).

[0014] Furthermore, when expressing a network diagram, the terms node and edge may be used. A node refers to a node or vertex in a network diagram, and corresponds to a document or community in this disclosure. An edge refers to a branch or edge in a network diagram, and corresponds to a link in this disclosure.

[0015] [First embodiment] FIG. 1 is a block diagram showing an example of an analysis device 1 according to the first embodiment. As shown in FIG. 1, the analysis device 1 includes an input unit 11, a display unit 12, a network communication unit 13, a storage unit 14, and a control unit 15. The analysis device 1 is a device that analyzes the relationships between a plurality of documents, and in this embodiment, a case where patent documents are used as an example of documents will be described.

[0016] The input unit 11 is an input device such as a keyboard and a mouse, and receives various input information for operating the analysis device 1. The display unit 12 is, for example, a display device such as a liquid crystal display, and displays the operation screen and analysis results of the analysis device 1. The display unit 12 displays, for example, a network diagram, which will be described later.

[0017] The NW (network) communication unit 13 is a communication module such as a wired LAN (Local Area Network) or a wireless LAN, and connects to a communication network to perform various data communications via the communication network. The NW communication unit 13 acquires documents such as patent documents and related information via the communication network.

[0018] The storage unit 14 stores various types of information used by the analysis device 1. The storage unit 14 includes, for example, a target list storage unit 141, a related information storage unit 142, an edge list storage unit 143, a node list storage unit 144, and a network storage unit 145.

[0019] The target list storage unit 141 stores a list of patents to be analyzed (a list of patent documents). The target list storage unit 141 stores a patent list that associates, for example, in-house identification information of a patent, patent application number, filing date, invention name, and examination status. The patent list to be analyzed stored in the target list storage unit 141 may be obtained, for example, from an external storage device (not shown) or via a communication network, or may be stored in the target list storage unit 141 in advance.

[0020] The related information storage unit 142 stores related information corresponding to the patent to be analyzed. The related information includes, for example, cited document information (application number of the cited patent), cited document information (application number of the cited patent), attribute information (applicant information, filing date, information on whether or not a patent right exists, etc.), related application information (information on divisional applications, converted applications, continuation applications, etc.), etc. The related information storage unit 142 stores the identification information (e.g., application number) of the patent to be analyzed in association with the related information.

[0021] The edge list storage unit 143 stores link information for each patent document and link information for each community. The link information is connection information having a direction for connecting nodes, and is information that associates node information of the "From" node (for example, an application number or identification information of a community) that is the source of the connection with node information of the "To" node (for example, an application number or identification information of a community) that is the source of the connection.

[0022] In this embodiment, the link direction is, for example, from cited documents to patent documents, from patent documents to cited documents, and from related child applications to parent applications (patent documents). Cited documents and cited documents are patent documents extracted from prior art documents of patent documents, cited documents in office action notices, etc. Related child applications are patent documents such as divisional applications, converted applications, or continuation applications, and parent applications are patent documents of the basic applications.

[0023] The node list storage unit 144 stores a list of node information (node ​​list) used when generating a network diagram, which will be described later. The node list includes a node list for each patent document and a node list for each community. The node information for each community may include the number of patent documents belonging to the community, keywords indicating the characteristics of the community (related keywords), and the like.

[0024] The network storage unit 145 stores data of the network diagram generated by the analysis device 1 (for example, image data, etc.).

[0025] The control unit 15 is, for example, a processor including a CPU (Central Processing Unit), and performs overall control of the analysis device 1. The control unit 15 executes control processing according to, for example, input information received from the input unit 11. The control unit 15 also acquires, for example, a patent list (list of patent documents) to be analyzed from outside the analysis device 1, and stores the patent list to be analyzed in the target list storage unit 141. The control unit 15 includes, for example, a link combining unit 151, a community analyzing unit 152, a network generating unit 153, and an output control unit 154.

[0026] The link combining unit 151 combines patent documents with links based on the citing / cited relationships of patent documents in a document group containing multiple documents, and generates an edge list showing a list of links for each patent document. The link combining unit 151 acquires related information, for example, via the NW communication unit 13 and a communication network, based on the patent list to be analyzed stored in the target list storage unit 141, and associates the patent to be analyzed with related information including cited document information and cited document information, and stores the associated information in the related information storage unit 142.

[0027] Furthermore, the link combining unit 151 generates an edge list for each patent document based on the related information (for example, relationship information between citing and cited patent documents) stored in the related information storage unit 142. For example, the link combining unit 151 extracts cited documents and cited documents from the related information stored in the related information storage unit 142, and generates an edge list of cited documents, cited documents, and patent documents.

[0028] The link combining unit 151 stores the generated edge list for each patent document in the edge list storage unit 143. The link combining unit 151 also generates a node list for each patent document and stores it in the node list storage unit 144. Note that the nodes for each patent document correspond to patent documents, and the link combining unit 151 stores the patent application numbers as identification information for the patent documents.

[0029] The community analysis unit 152 performs network analysis on the edge list for each document generated by the link connection unit 151, thereby dividing the target cluster (group of documents) into communities, which are groups of documents at lower levels, based on the density of the links. Here, network analysis refers to analyzing the properties of huge, complex networks that exist in the real world. Networks that exist in the real world are diverse and have huge, complex structures, but certain common properties can be identified. These properties are called "scale-freeness" (power law of degree distribution), "small-worldness," and "clustering" (quoted from the University of Tokyo Global Consumption Intelligence Endowed Chair (https: / / gci.tu-tokyo.ac.jp / tutorial / network / )). The community analysis unit 152 first performs cluster analysis using the edge list for each patent document stored in the edge list storage unit 143, and divides the patent list to be analyzed into clusters. The community analysis unit 152 then generates an edge list for each patent document for each cluster.

[0030] Furthermore, the community analysis unit 152 takes the designated cluster as the cluster to be analyzed, extracts communities that are sub-clusters from the edge list of patent documents that belong to the cluster to be analyzed, and divides the cluster to be analyzed into communities. For example, as shown in Figure 2, the community analysis unit 152 divides the cluster to be analyzed into multiple communities by using areas where the link density is low as the boundary of the community.

[0031] If the patent list to be analyzed contains multiple clusters, the community analysis unit 152 performs the same procedure for all clusters. The community analysis unit 152 also identifies an edge list for each community from the links for each document that connect communities. In this way, the community analysis unit 152 performs network analysis on the document-based edge list generated by the link connection unit 151, thereby dividing multiple patent documents into communities, which are document groups at a lower level of the document group (e.g., cluster), with boundaries defined by areas where the density of links is low, and identifies community-based edge lists from the document-based links that connect the communities.

[0032] FIG. 2 is a diagram showing an example of a network diagram for patent documents and division of communities in this embodiment. 2, the network diagram shows a network of patent documents in cluster CL1 to be analyzed. Each of communities CM1 to CM4 represents a divided community. As shown in FIG. 2, the community analysis unit 152 divides the network at the points where there is only one link (points with few links). The community analysis unit 152 executes the above-mentioned cluster and community extraction process using, for example, statistical analysis software "R." "R" provides many functions, such as the clusters function for cluster extraction, and walktrap.community (a community detection method based on a random walk), fastgreedy.community (a community detection method based on a greedy algorithm), and multilevel.community (a community detection method based on multilevel optimization), for community extraction. By selecting an applicable method from these various community detection methods for the cluster to be analyzed, the community division pattern can be uniquely determined.

[0033] The community analysis unit 152 also generates an edge list for each extracted community and stores the generated edge list in the edge list storage unit 143. The community analysis unit 152 also generates a node list for each community and stores the generated node list in the node list storage unit 144.

[0034] Here, the method for generating a community-based edge list will be explained below. If documents a and b are connected by a link and are classified into community A and community B, respectively, communities A and B are considered to be connected by a link. Also, if documents a and b are connected by a link and are both classified into community A, this is considered to be a self-reference from community A to community A and is ignored. By performing this procedure for all links listed in the document-based edge list, a community-based edge list will be obtained. The community-based edge list obtained by the above procedure may contain duplicate links. The duplicates may be left as they are, or the duplicate links may be deleted. If a duplicate link is deleted, a weight may be set for the link depending on the number of duplicates.

[0035] Returning to the explanation of Figure 1, the network generation unit 153 generates a network diagram for each community, which is illustrated by nodes indicating the community of the cluster (group of documents) and directional links based on an edge list for each community. The network generation unit 153 generates a network diagram based on the edge list stored in the edge list storage unit 143 and the node list stored in the node list storage unit 144. The network generation unit 153 generates a network diagram for each community, which is illustrated by nodes indicating the community of the cluster (group of documents) to be analyzed and directional links based on the citation and cited relationships between communities, for example.

[0036] Here, the network diagram is an example of output information illustrating the relationships between communities in the cluster to be analyzed. The network generation unit 153 generates, for example, network diagrams such as those shown in FIGS.

[0037] Fig. 3 is a diagram showing an example of a network diagram in units of communities in this embodiment, and Fig. 4 is a diagram showing an example of a network diagram in units of communities using a minimum spanning tree in this embodiment. 3 and 4, nodes represent communities, and the size of the circular figures (an example of an output figure) representing the nodes is changed according to the number of patent documents included in the community. That is, the network generation unit 153 generates a network diagram by changing the size of the circular figures (an example of an output figure) representing the nodes according to the number of patent documents included in the community. For example, the network generation unit 153 illustrates larger circular figures as the number of patent documents included in the community increases.

[0038] On the other hand, in Figures 3 and 4, the duplicated links have been deleted for edges. By deleting the duplicated links, a network diagram that is easier to read can be obtained. Conversely, by leaving the duplicated links or setting weights, the relationships between communities can be shown in more detail. The weights of these links are not shown in Figures 3 and 4. There are various ways to illustrate the weights, such as expressing the weights numerically, expressing the weights by the thickness of the lines, or expressing the weights by the color or shape of the lines.

[0039] In addition, in FIGS. 3 and 4, the numbers added to the nodes are identification information for identifying the nodes. The network generation unit 153 stores the generated network diagram for each community in the network storage unit 145 . The network generation unit 153 generates a network diagram for each community by processing the edge list for each community stored in the edge list storage unit 143 using the network diagram generation software "Cytoscape."

[0040] Returning to the explanation of FIG. 1 again, the output control unit 154 outputs output information that illustrates the relationships between communities in the target document collection, generated based on the communities. That is, the output control unit 154 outputs output information that illustrates the relationships between communities based on the communities and the edge list for each community. The output control unit 154 outputs, for example, a network diagram for each community generated by the network generation unit 153 to the display unit 12 as output information. That is, the output control unit 154 causes the network diagram for each community to be displayed on the display unit 12. Note that the output control unit 154 may output the network diagram for each community to a printer instead of the display unit 12, or may output it to the memory unit 14 or an external storage device and store it in a file.

[0041] The output control unit 154 also outputs a network diagram for each community using a minimum spanning tree, as shown in Fig. 4. The output control unit 154 generates a minimum spanning tree from the edge list for each community stored in the edge list storage unit 143, and causes the network generation unit 153 to generate a network diagram for each community using the minimum spanning tree. The output control unit 154 causes the display unit 12 to display the network diagram for each community using the minimum spanning tree as shown in Fig. 4, generated by the network generation unit 153.

[0042] Next, the operation of the analyzer 1 according to this embodiment will be described with reference to the drawings. FIG. 5 is a flowchart showing an example of a community generation process of the analysis device 1 according to this embodiment.

[0043] 5, the analysis device 1 acquires a list of target documents (step S101). The control unit 15 of the analysis device 1 acquires the list of patent documents to be analyzed, for example, from outside the analysis device 1. The control unit 15 stores the acquired patent document list in the target list storage unit 141.

[0044] Next, the control unit 15 acquires related information (step S102). The link connection unit 151 of the control unit 15 acquires the related information, for example, via the NW communication unit 13 and a communication network, based on the patent list to be analyzed stored in the target list storage unit 141. Here, the related information includes, for example, cited document information (application numbers of cited patents), cited document information (application numbers of cited patents), attribute information (applicant information, filing date, information on the existence of patent rights, etc.), related application information (application numbers of divisional applications, converted applications, continuation applications, etc.), etc. The link connection unit 151 associates identification information (e.g., application numbers) of the patent documents to be analyzed with the related information and stores the associated information in the related information storage unit 142.

[0045] Next, the link combining unit 151 combines the documents with links based on the citing / cited relationships of each document, and generates a node list and an edge list for each document (step S103). The link combining unit 151 extracts link information that combines patent documents, for example, by linking the application number of the analysis target with the application number of the cited document, linking the application number of the analysis target with the application number of the cited document, and linking the application number of the analysis target with the application number of the related application. The link combining unit 151 associates the extracted link information with the application number of the link source and the application number of the link destination, and stores the extracted link information in the edge list storage unit 143. The link combining unit 151 also generates a node list for each patent document, and stores the generated node list in the node list storage unit 144.

[0046] Next, the community analysis unit 152 of the control unit 15 extracts clusters and generates an edge list for each cluster (step S104). The community analysis unit 152 extracts clusters from the edge list for each patent document stored in the edge list storage unit 143, for example, using statistical analysis software "R." "R" provides a function "clusters" for cluster extraction. The community analysis unit 152 generates an edge list for each patent document for each extracted cluster.

[0047] Next, the community analysis unit 152 may perform pruning processing. For example, the community analysis unit 152 performs processing to reduce the number of nodes in the edge list for each patent document for each cluster by deleting all patent documents other than the patent document to be analyzed and patent documents connecting the patent documents to be analyzed. By performing pruning processing in this way, it is possible to make it easier to understand the relationships between the patent documents to be analyzed.

[0048] Next, the community analysis unit 152 performs community analysis of the target cluster and generates an edge list for each community (step S105). The community analysis unit 152, for example, sets a cluster specified via the input unit 11 as the cluster to be analyzed, and extracts communities, which are sub-clusters, from the edge list of the cluster to be analyzed using, for example, statistical analysis software "R." "R" provides many functions for community extraction, such as walktrap.community, fastgreedy.community, and multilevel.community. The community analysis unit 152 generates an edge list for each extracted community. The community analysis unit 152 stores the edge list for each community in the edge list storage unit 143. After processing in step S105, the community analysis unit 152 ends the community generation process.

[0049] Next, the process of generating a network diagram for each community by the analysis device 1 according to this embodiment will be described with reference to FIG. FIG. 6 is a flowchart showing an example of a process for generating a network diagram for each community by the analysis device 1 according to this embodiment. The generation process described in FIG. 6 is for outputting a network diagram for each community using a minimum spanning tree, as shown in FIG. Of course, it is also possible to create a network diagram that illustrates all the relationships between communities, as shown in FIG. 3, but a description of this will be omitted here.

[0050] As shown in FIG. 6, the output control unit 154 of the analysis device 1 generates an edge list of a minimum spanning tree from the edge list of each community of the target cluster (step S201). The output control unit 154 acquires the edge list of each community of the target cluster from the edge list storage unit 143, and generates an edge list of a minimum spanning tree from the edge list of each community using a method such as the Kruskal algorithm. In addition, "R" provides a function minimum.spanning.tree for obtaining a minimum spanning tree. The output control unit 154 stores the generated edge list of the minimum spanning tree in the edge list storage unit 143 as an edge list of each community.

[0051] Next, the network generation unit 153 of the analysis device 1 generates a network diagram from the edge list and attribute information of the minimum spanning tree for each community (step S202). Here, the output control unit 154 requests the network generation unit 153 to perform processing to generate a network diagram, and the network generation unit 153 executes the processing to generate a network diagram.

[0052] The network generation unit 153 extracts community nodes from the community-based edge list stored in the edge list storage unit 143, and generates a node list that associates node identification information with attribute information such as the number of patent documents belonging to the community. The network generation unit 153 generates a network diagram based on the community-based edge list and node list, for example, using network diagram generation software "Cytoscape."

[0053] In addition, the network generation unit 153 changes the size of the circular figures indicating the nodes in the network diagram according to the number of patent documents belonging to the community. The network generation unit 153 stores the generated network diagram in the network storage unit 145.

[0054] Next, the output control unit 154 outputs the network diagram generated by the network generation unit 153 to the display unit 12 (step S203). The output control unit 154 causes the display unit 12 to display the network diagram stored in the network storage unit 145. For example, the output control unit 154 causes the display unit 12 to display a network diagram of a minimum spanning tree as shown in Fig. 4. After the processing of step S203, the output control unit 154 ends the process of generating a network diagram for each community.

[0055] As shown in FIG. 7, the output control unit 154 may add keywords to the network diagram of the minimum spanning tree for each community and display it on the display unit 12. In this case, the output control unit 154 analyzes patent documents belonging to each community in response to instruction information for adding keywords received from the input unit 11, and extracts keywords that characterize each community. The output control unit 154 extracts keywords from the contents of the "title of the invention," "abstract," and "description" in the patent documents, for example, using text mining software such as "KH Coder." The output control unit 154 displays the extracted keywords superimposed on the node position of each community. Here, the keywords are preferably words described in patent documents belonging to a community that appear frequently in that community and infrequently in other communities.

[0056] As described above, the analysis device 1 according to this embodiment includes a link combining unit 151, a community analysis unit 152, and an output control unit 154. The link combining unit 151 combines patent documents with links based on the citing and cited relationships of patent documents in a document group (e.g., a cluster) containing multiple documents, and generates an edge list showing a list of links for each patent document. The community analysis unit 152 performs network analysis on the document-by-document edge list generated by the link combining unit 151 to divide the multiple documents into communities, which are document groups at a lower level than the document group (e.g., a cluster), with boundaries defined by areas where the links are sparse. The output control unit 154 then outputs output information illustrating the relationships between the communities based on the communities and the community-by-community edge list. Here, the documents are, for example, patent documents.

[0057] As a result, the analysis device 1 according to this embodiment can visually grasp the citing and cited relationships between communities that include a group of technically related documents (e.g., patent documents), allowing for an overview of the relationships between patent groups and easy identification of noteworthy communities (patent groups). In other words, the analysis device 1 according to this embodiment can appropriately grasp the interrelationships between documents (e.g., patent documents) and the characteristics of a group of documents (patent document groups).

[0058] The analysis device 1 according to this embodiment also includes a network generation unit 153 that generates a network diagram for each community illustrated with nodes indicating communities of a group of target documents and directional links based on citation-citation relationships between the communities. The network generation unit 153 generates a network diagram for each community illustrated with nodes indicating communities of clusters (groups of documents) and directional links based on an edge list for each community. The output control unit 154 outputs the network diagram for each community generated by the network generation unit 153 as output information.

[0059] As a result, the analysis device 1 according to this embodiment can visually grasp the citation and citation relationships between communities by using a network diagram on a community-by-community basis, making it possible to get an overview of the relationships between groups of patents and easily identify noteworthy communities (groups of patents).

[0060] For example, for comparison, Fig. 8 shows an example of a network diagram for patent documents. In a network diagram for patent documents such as that shown in Fig. 8, as the number of documents increases, the network diagram becomes complicated, making it difficult to properly grasp the interrelationships between patent documents and the characteristics of the document group.

[0061] 9 is a diagram showing an example of a network diagram in units of communities in this embodiment. The network diagram in units of communities shown in FIG. 9 is a network diagram for the same cluster to be analyzed as in FIG. As shown in FIG. 9, the network diagram for each community in this embodiment is a simplified network diagram compared to FIG. 8, making it easier to grasp the interrelationships between nodes (communities) and the characteristics of the clusters being analyzed.

[0062] In this embodiment, the output control unit 154 outputs a network diagram for each community using a minimum spanning tree. As a result, the analysis device 1 according to this embodiment connects communities with only the links with the highest weights, as shown in FIG. 4, for example, and therefore makes it easier to grasp the relationships between communities.

[0063] Furthermore, in this embodiment, the output control unit 154 changes the size of the output graphic (for example, a circular graphic) showing the node in the network diagram for each community, depending on the number of documents included in the community. As a result, the analysis device 1 according to this embodiment can easily grasp the size of the community, and can therefore more easily identify noteworthy communities (patent groups).

[0064] In this embodiment, the documents are patent documents. The link combining unit 151 combines documents by links based on the citing / cited relationships of the patent documents and information on divisional applications, converted applications, and continuation applications of the patent documents. As a result, the analysis device 1 of this embodiment analyzes the relationships between patent documents, including divisional applications, converted applications, and continuation applications of patent documents, thereby making it possible to more appropriately grasp the interrelationships between patent documents and the characteristics of groups of patent documents.

[0065] In addition, in this embodiment, the output control unit 154 extracts keywords that indicate the characteristics of the community from the contents of the documents belonging to the community, and outputs a network diagram (output information) with the keywords added to a position corresponding to the community. This makes it easier for the analysis device 1 according to this embodiment to grasp the characteristics of the community, thereby improving convenience.

[0066] In this embodiment, the output control unit 154 causes the display unit 12 to output output information (for example, a network diagram). As a result, the analyzer 1 according to this embodiment makes it possible to easily check the output information (for example, a network diagram) displayed on the display unit 12.

[0067] The analysis method according to this embodiment is an analysis method using the above-described analysis device 1, and includes a link combining step, a community analysis step, and an output control step. In the link combining step, the link combining unit 151 combines patent documents with links based on the citing / cited relationships of patent documents in a document group (e.g., a cluster) including multiple documents, and generates an edge list showing a list of links for each patent document. In the community analysis step, the community analysis unit 152 performs network analysis on the document-by-document edge list generated by the link combining unit 151, thereby dividing the multiple documents into communities, which are document groups at a lower level than the document group (e.g., a cluster), with boundaries defined by areas where the links are sparse, and identifies edge lists for each community from the document-by-document links connecting the communities. In the output control step, the output control unit 154 outputs output information illustrating the relationships between the communities based on the communities and the edge lists for each community.

[0068] As a result, the analysis method according to this embodiment has the same effect as the above-described analysis device 1, and can appropriately grasp the interrelationships between documents (for example, patent documents) and the characteristics of a document group (patent document group).

[0069] [Second embodiment] Next, an analysis device 1a according to a second embodiment will be described with reference to the drawings. FIG. 10 is a block diagram showing an example of an analysis device 1a according to the second embodiment. As shown in FIG. 10, the analysis device 1a includes an input unit 11, a display unit 12, a NW communication unit 13, a storage unit 14a, and a control unit 15a.

[0070] In this embodiment, a modified example will be described in which the analysis device 1a outputs a heat map diagram as output information in addition to a network diagram. In FIG. 10, the same components as those in the first embodiment shown in FIG. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0071] The storage unit 14a stores various types of information used by the analysis device 1a. The storage unit 14a includes, for example, a target list storage unit 141, a related information storage unit 142, an edge list storage unit 143, a node list storage unit 144, a network storage unit 145, and a heat map storage unit 146.

[0072] The heat map storage unit 146 stores data (for example, image data) of the heat map diagram generated by the analysis device 1a.

[0073] The control unit 15a is, for example, a processor including a CPU, and performs overall control of the analysis device 1a. The control unit 15a executes control processing according to, for example, input information received from the input unit 11. The control unit 15a also acquires, for example, a list of patents to be analyzed (a list of patent documents) from outside the analysis device 1a, and stores the list of patents to be analyzed in the target list storage unit 141. The control unit 15a includes, for example, a link combining unit 151, a community analyzing unit 152, a network generating unit 153, an output control unit 154a, and a heat map generating unit 155.

[0074] The heat map generation unit 155 generates a heat map diagram for each community based on the number of patent documents included in the community of the cluster (group of documents) to be analyzed and the edge list for each community. The heat map generation unit 155 calculates a distance matrix indicating the similarity between communities based on the edge list for each community and the number of patent documents belonging to the community. Here, the distance matrix is ​​based on, for example, the Jaccard coefficient.

[0075] The heat map generation unit 155 calculates the two-dimensional position coordinates of the communities, for example, using multidimensional scaling (MDS), and arranges each community two-dimensionally at the calculated position. Here, the higher the similarity between the communities, the closer they are arranged, and the lower the similarity between the communities, the farther they are arranged. The heat map generation unit 155 converts the number of patent documents included in the community into a heat map, for example, using kernel density estimation, and generates a heat map diagram for each community. Note that in the heat map diagram for each community, the similarity between communities is illustrated as the distance between the communities.

[0076] The heat map generating unit 155 stores the generated heat map diagram for each community in the heat map storage unit 146.

[0077] The output control unit 154a outputs, as output information, the heat map diagram for each community generated by the heat map generation unit 155. For example, the output control unit 154a causes the display unit 12 to display a heat map diagram for each community as shown in FIG.

[0078] 11 is a heat map diagram with keywords added. In response to instruction information for adding keywords received from the input unit 11, the output control unit 154a displays the keywords on the display unit 12, superimposing them on the positions of the nodes of each community. In addition, in the heat map shown in FIG. 11, the numbers in the map are identification numbers (identification information) that identify the communities.

[0079] Furthermore, the output control unit 154a outputs output information in which a network diagram for each community is superimposed on a heat map diagram for each community. In response to instruction information for adding a network diagram received from the input unit 11, the output control unit 154a causes the display unit 12 to display the network diagram for each community generated by the network generation unit 153 superimposed on the heat map diagram for each community. For example, as shown in FIG. 12, the output control unit 154a causes the display unit 12 to display the network diagram superimposed on the heat map diagram.

[0080] FIG. 12 is a diagram showing an example of a case where a network diagram for each community is superimposed on a heat map diagram for each community in this embodiment. As shown in FIG. 12, the output control unit 154a causes the display unit 12 to display a network diagram superimposed on the heat map diagram so that the position indicating the community in the community-based heat map diagram matches the position of the community node in the network diagram.

[0081] Next, the operation of the analysis device 1a according to this embodiment will be described with reference to the drawings. The operation described here is for overlaying a network diagram of a minimum spanning tree on a heat map diagram and displaying it on the display unit 12, as shown in FIG. The community generation process of the analysis device 1a according to this embodiment is similar to the process of the first embodiment shown in FIG. 6, and therefore will not be described here.

[0082] FIG. 13 is a flowchart showing an example of a process for generating a heat map diagram for each community by the analysis device 1a according to this embodiment. In FIG. 13, the processes of steps S301 and S302 are the same as the processes of steps S201 and S202 shown in FIG. 6 described above, and therefore a description thereof will be omitted here.

[0083] In step S303, the heat map generation unit 155 of the analysis device 1a generates a heat map from the edge list and node list for each community. Here, the output control unit 154a requests the heat map generation unit 155 to perform processing to generate a heat map, and the heat map generation unit 155 executes the processing to generate a heat map.

[0084] The heat map generation unit 155 acquires the number of patent documents belonging to each community, for example, from a node list. The heat map generation unit 155 generates a heat map diagram for each community based on the edge list for each community and the number of patent documents belonging to the community. The heat map generation unit 155 stores the generated heat map diagram for each community in the heat map storage unit 146.

[0085] Next, the output control unit 154a outputs the network diagram generated by the network generation unit 153 and the heat map diagram generated by the heat map generation unit 155 to the display unit 12 (step S304). For example, the output control unit 154a causes the display unit 12 to display a network diagram of the minimum spanning tree for each community superimposed on the heat map diagram for each community stored in the heat map storage unit 146, as shown in Fig. 12. After the processing of step S304, the output control unit 154a ends the processing.

[0086] As described above, the analysis device 1a according to this embodiment includes the heat map generation unit 155 that generates a heat map diagram for each community based on the number of documents included in the community of the target document group and the edge list for each community. The output control unit 154a outputs the heat map diagram for each community generated by the heat map generation unit 155 as output information.

[0087] As a result, the analysis device 1a according to this embodiment can easily grasp a community that includes a group of patent documents that are highly technically related, and can visually grasp the number of patent documents included in the community, using a heat map diagram for each community. Therefore, the analysis device 1a according to this embodiment can easily identify a community that requires attention.

[0088] For example, for comparison, Figure 14 shows an example of a heat map diagram by patent document. Each dot in the diagram represents one patent document. In a heat map diagram by patent document such as that shown in Figure 14, when the number of documents increases, the display of patent documents becomes too detailed (too small), making it difficult to properly grasp the interrelationships between patent documents and the characteristics of the document group.

[0089] 15 is a diagram showing an example of a heat map diagram in units of communities in this embodiment. The heat map diagram in units of communities shown in FIG. 15 is a heat map diagram for the same cluster to be analyzed as in FIG. As shown in Figure 15, the heat map diagram per community in this embodiment is a simplified heat map diagram compared to Figure 14, making it easier to understand the interrelationships between nodes (communities) and the characteristics of the cluster being analyzed.

[0090] The analysis device 1a according to this embodiment also includes a network generation unit 153 that generates a network diagram for each community, which is illustrated with nodes indicating communities of a group of target documents and directional links based on citation-citation relationships between the communities. The network generation unit 153 generates a network diagram for each community, which is illustrated with nodes indicating communities of clusters (groups of documents) and directional links based on an edge list for each community. The output control unit 154a outputs output information (see, for example, FIG. 12) in which a network diagram for each community is superimposed on a heat map diagram for each community.

[0091] As a result, the analysis device 1a according to this embodiment can more easily grasp the relationships between communities by overlaying a network diagram for each community on a heat map diagram for each community.

[0092] Furthermore, in this embodiment, the output control unit 154a outputs the output information (e.g., a heat map diagram and a network diagram) by illustrating the similarity between communities by the distance between the communities. As a result, the analysis device 1a according to this embodiment can more intuitively grasp the relationship between communities based on the distance between the communities.

[0093] Next, a display modification and a comparative example of the heat map diagram of the analysis device 1a according to this embodiment will be described with reference to FIGS. In this embodiment, the output control unit 154a may output output information for the patent document corresponding to the specified attribute information.

[0094] For example, the attribute information includes information on whether a patent right exists. Here, the information on whether a patent right exists is information indicating whether the patent right is valid (still valid) or invalid (lost). For example, the output control unit 154a may display, on the display unit 12, a heat map diagram for each community generated from a group of patent documents for which the patent right is valid (still valid) and a heat map diagram for each community generated from a group of patent documents for which the patent right is invalid (lost), so that these can be compared.

[0095] Fig. 16 is a diagram showing a first comparative example of a heat map diagram by community unit in this embodiment. Fig. 16(a) shows a heat map diagram by community unit for all patent documents. Fig. 16(b) shows a heat map diagram by community unit for patent documents whose patent rights are valid (still in force). Fig. 16(c) shows a heat map diagram by community unit for patent documents whose patent rights are invalid (expired).

[0096] In this case, the output control unit 154a may display any one of the heat maps shown in Figures 16(a) to 16(c) in accordance with the display specification information specified via the input unit 11. Furthermore, the output control unit 154a may display the heat maps shown in Figures 16(a) to 16(c) side by side for comparison in accordance with the display specification information specified via the input unit 11.

[0097] As a result, the analysis device 1a according to this embodiment can appropriately grasp the current status of patent documents by outputting output information for the patent documents corresponding to the specified information on the existence of patent rights.

[0098] The attribute information also includes, for example, information about the applicant. Here, the applicant information is information indicating, for example, whether the application is filed by the company or another company. The output control unit 154a may, for example, display on the display unit 12 a heat map diagram for each community generated from a group of patent documents of the company's applications and a heat map diagram for each community generated from a group of patent documents of other companies' applications, so that these can be compared.

[0099] Fig. 17 shows a second comparative example of a heat map diagram by community in this embodiment. Fig. 17(a) shows a heat map diagram by community based on patent documents filed by the company itself. Fig. 17(b) shows a heat map diagram by community based on patent documents filed by other companies.

[0100] In this case, the output control unit 154a may display the heat map diagram of Fig. 17(a) or Fig. 17(b) in accordance with the display specification information specified via the input unit 11. Furthermore, the output control unit 154a may display the heat map diagrams of Fig. 17(a) and Fig. 17(b) side by side for comparison in accordance with the display specification information specified via the input unit 11.

[0101] As a result, the analysis device 1a according to this embodiment can appropriately grasp the differences between the current state of the patent document of the company and the current state of the patent document of another company by outputting output information for the patent document corresponding to the specified company or another company.

[0102] The attribute information may also include, for example, the filing date of the patent. The output control unit 154a may, for example, display on the display unit 12 a heat map diagram for each community generated from a group of patent documents filed before a specified date, allowing comparison of these by era.

[0103] Fig. 18 is a diagram showing a third comparative example of a heat map diagram per community in this embodiment. Fig. 18(a) shows a heat map diagram per community based on patent documents filed between 2000 and 2004. Fig. 18(b) shows a heat map diagram per community based on patent documents filed between 2012 and 2016. Fig. 18(c) shows a heat map diagram per community based on patent documents filed between 2016 and 2020.

[0104] In this case, the output control unit 154a may display heat maps by era as shown in Figures 18(a) to 18(c) in accordance with the display designation information (information on the application year) designated via the input unit 11. Furthermore, the output control unit 154a may display the heat maps of Figures 18(a) to 18(c) side by side for comparison in accordance with the display designation information (information on the application year) designated via the input unit 11.

[0105] As a result, the analysis device 1a according to this embodiment can appropriately grasp the change over time in the application trends of patent documents by outputting output information for patent documents corresponding to the specified range of application dates.

[0106] The present disclosure is not limited to the above-described embodiments, and can be modified within the scope of the present disclosure. For example, in each of the above embodiments, the analytical device 1 (1a) is described as being configured as a single device, but this is not limited thereto and the analytical device 1 (1a) may be configured as being configured as a plurality of devices. For example, the analytical device 1 (1a) may be configured such that a part of the control unit 15 (15a) and the storage unit 14 (14a) is provided in an external device (for example, a server device connected via a communication network).

[0107] Furthermore, in each of the above embodiments, examples have been described in which the documents to be analyzed are patent documents, but this is not limited to this, and the documents to be analyzed may be any documents that are associated with citations and cited documents, and may be, for example, other technical documents such as academic papers.

[0108] Furthermore, the attribute information associated with a patent document is not limited to that described in the above embodiments, but may be other information such as the International Patent Classification (IPC). In addition, in the above second embodiment, an example was described in which the similarity between communities is illustrated by the distance between the communities and output information (e.g., a heat map diagram and a network diagram), but this may also be applied to the first embodiment.

[0109] In each of the above embodiments, the analysis device 1 (1a) may be configured to allow a user to specify a set of patent documents to be included in the same community so that patent documents known to have a strong relationship in advance are not separated into separate communities. In this case, the community analysis unit 152 prevents a set of patent documents specified to be included in the same community via the input unit 11 from being separated when extracting the community. In the calculation process, a set of patent documents specified to be included in a community may be processed as a single patent document.

[0110] In each of the above embodiments, the analysis device 1 (1a) may change the strength of the link depending on the type of cited document, for example. For example, a weight may be assigned to each link, so that the patent document that is the primary cited document in the notice of reasons for refusal is weighted more heavily than the other patent documents.

[0111] Each component of the analytical device 1 (1a) described above has an internal computer system. A program for implementing the functions of each component of the analytical device 1 (1a) described above may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed to perform processing in each component of the analytical device 1 (1a). Here, "reading a program recorded on a recording medium into a computer system and executing it" includes installing the program into a computer system. The term "computer system" here includes hardware such as an OS and peripheral devices.

[0112] Furthermore, a "computer system" may include multiple computer devices connected via a network, including communication lines such as the Internet, WAN, LAN, and dedicated lines. Furthermore, a "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into a computer system. Thus, the recording medium storing the program may be a non-transitory recording medium such as a CD-ROM.

[0113] The recording medium also includes internal or external recording media accessible from a distribution server for distributing the program. The program may be divided into multiple parts, downloaded at different times, and then combined into each component of the analysis device 1 (1a), or each divided program may be distributed by a different distribution server. Furthermore, the term "computer-readable recording medium" also includes a storage medium that stores a program for a certain period of time, such as volatile memory (RAM) within a computer system that serves as a server or client when a program is transmitted over a network. The program may also be a storage medium for implementing part of the above-described functions. Furthermore, the program may be a so-called differential file (differential program) that can realize the above-described functions in combination with a program already stored in the computer system. [Explanation of symbols]

[0114] 1, 1a... analysis device, 11... input unit, 12... display unit, 13... network communication unit, 14, 14a... storage unit, 15, 15a... control unit, 141... target list storage unit, 142... related information storage unit, 143... edge list storage unit, 144... node list storage unit, 145... network storage unit, 146... heat map storage unit, 151... link connection unit, 152... community analysis unit, 153... network generation unit, 154, 154a... output control unit, 155... heat map generation unit

Claims

1. a link combining unit that combines documents with links based on citing / cited relationships in a document group including a plurality of documents, and generates an edge list showing a list of links for each document; a community analysis unit that performs a network analysis on the edge list per document unit generated by the link combination unit to divide the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are areas where the links are sparse, and identifies the edge list per community unit from the document links that connect the communities; an output control unit that outputs output information illustrating relationships between the communities based on the communities and the edge list for each community; a heat map generating unit that generates a heat map diagram for each community based on the number of documents included in the community of the document group and an edge list for each community; a network generation unit that generates a network diagram for each community illustrated by nodes indicating the community of the document group and directional links based on an edge list for each community; Equipped with The output control unit The output information is output in which the network diagram for each community generated by the network generation unit is superimposed on the heat map diagram for each community generated by the heat map generation unit. Analyzer.

2. a network generation unit that generates a network diagram for each community illustrated by nodes indicating the community of the document group and directional links based on an edge list for each community; The output control unit The network diagram for each community generated by the network generation unit is output as the output information. The analytical device of claim 1 .

3. The output control unit outputs the network diagram for each community using a minimum spanning tree. The analytical device according to claim 1 or 2.

4. The output control unit changes the size of an output graphic showing the node in the network diagram for each community in accordance with the number of documents included in the community. The analytical device according to claim 2 or 3.

5. The output control unit outputs the output information by illustrating the similarity between the communities by indicating the distance between the communities. The analysis device according to any one of claims 1 to 4.

6. A link combining unit that combines documents with links based on the citing / cited relationships of a document group containing multiple documents, and generates an edge list that shows a list of links for each document; a community analysis unit that performs a network analysis on the edge list per document unit generated by the link combination unit to divide the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are areas where the links are sparse, and identifies the edge list per community unit from the document links that connect the communities; an output control unit that outputs output information illustrating relationships between the communities based on the communities and the edge list for each community; Equipped with the document is a patent document, The link combining unit combines the documents by links based on the citing / cited relationship of the patent document and information on divisional applications, converted applications, and continuation applications of the patent document. Analyzer.

7. The output control unit extracts keywords indicating characteristics of the community from the description of the document belonging to the community, and outputs the output information to which the keywords are added at a position corresponding to the community. The analytical device according to any one of claims 1 to 6.

8. The output control unit outputs the output information for the document corresponding to the specified attribute information. The analytical device according to any one of claims 1 to 7.

9. The output control unit causes a display unit to output the output information. The analytical device according to any one of claims 1 to 8.

10. a link linking step in which a link linking unit links documents based on citing / cited relationships in a document group including a plurality of documents, and generates an edge list showing a list of links for each document; a community analysis step in which a community analysis unit performs a network analysis on the edge list per document unit generated by the link connection step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are the parts where the links are sparse, and identifying the edge list per community unit from the document links connecting the communities; an output control step in which an output control unit outputs output information illustrating relationships between the communities based on the communities and the edge list for each community; a heat map generating step in which a heat map generating unit generates a heat map diagram for each community based on the number of documents included in the community of the document group and an edge list for each community; a network generation step in which a network generation unit generates a network diagram for each community illustrated by nodes indicating the communities of the document group and directional links based on an edge list for each community; Including, In the output control step, the output control unit outputs the output information in which the network diagram for each community generated in the network generation step is superimposed on the heat map diagram for each community generated in the heat map generation step. Analysis method.

11. On the computer, a link linking step of linking documents based on citing / cited relationships in a document group including a plurality of documents and generating an edge list showing a list of links for each document; a community analysis step of performing a network analysis on the edge list per document unit generated by the link connection step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are the parts where the links are sparse, and identifying the edge list per community unit from the document links connecting the communities; an output control step of outputting output information illustrating relationships between the communities based on the communities and the edge list for each community; a heat map generating step in which a heat map generating unit generates a heat map diagram for each community based on the number of documents included in the community of the document group and an edge list for each community; a network generation step in which a network generation unit generates a network diagram for each community illustrated by nodes indicating the communities of the document group and directional links based on an edge list for each community; It is a program for executing In the output control step, the output information is output in which the network diagram for each community generated in the network generation step is superimposed on the heat map diagram for each community generated in the heat map generation step. program.

12. A link linking step in which a link linking unit links documents based on citing / cited relationships in a document group including multiple documents, and generates an edge list showing a list of links for each document; a community analysis step in which a community analysis unit performs a network analysis on the edge list per document unit generated by the link connection step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are the parts where the links are sparse, and identifying the edge list per community unit from the document links connecting the communities; an output control step in which an output control unit outputs output information illustrating relationships between the communities based on the communities and the edge list for each community; Including, the document is a patent document, In the linking step, the linking unit links the documents based on the citing / cited relationships of the patent documents and information on divisional applications, converted applications, and continuation applications of the patent documents. Analysis method.

13. A computer comprising: a link linking step of linking documents based on citing / cited relationships in a document group including a plurality of documents and generating an edge list showing a list of links for each document; a community analysis step of performing a network analysis on the edge list per document unit generated by the link connection step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are the parts where the links are sparse, and identifying the edge list per community unit from the document links connecting the communities; an output control step of outputting output information illustrating relationships between the communities based on the communities and the edge list for each community; It is a program for executing the document is a patent document, In the linking step, the documents are linked based on the citing / cited relationships of the patent documents and information on divisional applications, converted applications, and continuation applications of the patent documents. program.

14. A link combining unit that combines documents with links based on citing / cited relationships in a document group containing multiple documents, and generates an edge list that lists links for each document; a community analysis unit that performs a network analysis on the edge list per document unit generated by the link combination unit to divide the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are areas where the links are sparse, and identifies the edge list per community unit from the document links that connect the communities; an output control unit that outputs output information illustrating relationships between the communities based on the communities and the edge list for each community; a heat map generating unit that generates a heat map diagram for each community based on the number of documents included in the community of the document group and the edge list for each community; Equipped with The output control unit outputs the output information in which the heat map diagrams corresponding to each of a plurality of pieces of specified attribute information different from each other are displayed side by side so as to be comparable. Analyzer.

15. A link linking step in which a link linking unit links documents based on citing / cited relationships in a document group including multiple documents, and generates an edge list showing a list of links for each document; a community analysis step in which a community analysis unit performs a network analysis on the edge list per document unit generated by the link connection step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are the parts where the links are sparse, and identifying the edge list per community unit from the document links connecting the communities; an output control step in which an output control unit outputs output information illustrating relationships between the communities based on the communities and the edge list for each community; Including, In the output control step, the output control unit outputs the output information in which the heat map diagrams for each community corresponding to each of a plurality of pieces of specified attribute information different from each other are displayed side by side so as to be comparable. Analysis method.

16. A computer comprising: a link linking step of linking documents based on citing / cited relationships in a document group including a plurality of documents and generating an edge list showing a list of links for each document; a community analysis step of performing a network analysis on the edge list per document unit generated by the link connection step, thereby dividing the plurality of documents into communities, which are document groups at a lower level than the document group, so that the boundaries are the parts where the links are sparse, and identifying the edge list per community unit from the document links connecting the communities; an output control step of outputting output information illustrating relationships between the communities based on the communities and the edge list for each community; It is a program for executing In the output control step, the heat map diagrams for each community corresponding to each of a plurality of pieces of specified attribute information different from each other are output as output information in a juxtaposed manner so that they can be compared. program.

Images