Regional index evaluation device and regional index evaluation method

The regional index evaluation device addresses the challenge of interdependent regional indicators by creating an integrated graph to evaluate and optimize facility placement, ensuring balanced development and planning across multiple indicators.

JP2024157380APending Publication Date: 2024-11-07HITACHI LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2023071710
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-04-25
Publication Date
2024-11-07

AI Technical Summary

Technical Problem

Existing technologies fail to consider the interdependencies among regional indicators when deploying or relocating public and business services, leading to potential negative impacts on other indicators, such as increased environmental burden for improved economic efficiency.

Method used

A regional index evaluation device that evaluates the relationships among multiple regional indicators by creating an integrated graph with nodes representing regional indicators, facility types, and points, using machine learning to calculate connection strengths and feature vectors, thereby facilitating balanced regional development.

Benefits of technology

Enables balanced regional development and planning by evaluating the relationships among regional indicators, allowing for informed decisions on facility placement to optimize multiple indicators simultaneously.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024157380000001_ABST
    Figure 2024157380000001_ABST
Patent Text Reader

Abstract

To evaluate the relationship in a plurality of regional indices for a facility and provision of a service in a certain region, to allow balanced regional development and regional planning.SOLUTION: A regional index evaluation device holds an index and facility-related evaluation table storing a facility type related to indices, an evaluation value of the relationship between the indices, and an evaluation value of the relationship between the indices and the facility type, and a point information table storing an evaluation value of the relationship between points in a region and the facility type, on the basis of the tables, creates an integrated graph that has an edge with the indices, facility type, and points as nodes and the indices and the indices as nodes, an edge with the indices and facility type as nodes, an edge with the facility type and points as nodes, and has the weights on the respective edges, calculates a feature vector of each of the nodes of the integrated graph, calculates the coupling strength between the nodes from the feature vectors of nodes corresponding to the indices, and evaluates the strength of the relationship between the indices on the basis of the coupling strength between the nodes.SELECTED DRAWING: Figure 11
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a regional index evaluation device and a regional index evaluation method, and more particularly to a regional index evaluation device and a regional index evaluation method that are suitable for use in supporting the formulation of regional policies by local governments and the formulation of business plans by businesses by linking regional development with regional indexes. [Background technology]

[0002] In recent years, efforts have been made to introduce appropriate regional indicators as regional indicators that promote the realization of integrated improvements in the environment, economy, and society in the region, thereby aiming for balanced regional development. For example, Non-Patent Document 1 explains specific examples and their significance as evaluation indicators of the environment, economy, and society that take into account influences both within and outside the region.

[0003] Furthermore, technology for evaluating and calculating regional indices using an information processing device is, for example, found in Patent Document 1. In the regional index evaluation system described in Patent Document 1, the amount of power used and the amount of power consumed per unit of water distribution are given as regional indices (city indices), and technology is disclosed for grouping with other cities by performing correlation analysis with characteristics such as the average temperature and the elevation difference of the land (Fig. 12, paragraphs

[0090] -

[0099] ). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2016-206772 A [Non-patent literature]

[0005] [Non-Patent Document 1] Hideaki Kurishima and 8 others, "Development of evaluation indicators and measurement methods for the environment, economy, and society considering the impacts inside and outside the region", [online], Policy research on environmental economy for FY2014 commissioned by the Ministry of the Environment, [Retrieved April 12, 2023], Internet<URL:https: / / www.env.go.jp / policy / keizai_portal / F_research / 4Presen.pdf> Summary of the Invention [Problem to be solved by the invention]

[0006] In a community, there are public services such as government offices and community centers, and business services such as restaurants and hospitals. Previously, when trying to establish or relocate these services, it was difficult for the public and private sectors to know in advance what impact it would have on the respective regional indicators (resident satisfaction, economic effects, environment, etc.). In addition, it could happen that an improvement in one of the regional indicators results in the other indicators worsening (e.g., an increase in environmental impact as a result of a focus on improving economic efficiency). Conventional technology did not consider the technology that would allow the public and private sectors to consider such situations in advance and proceed with plans while balancing various indicators.

[0007] The technology described in Patent Document 1 evaluates multiple city indicators related to the amount of electricity used and the amount of electricity consumed per unit of water distribution by assigning a score to each city indicator and is able to recommend other similar cities. However, this technology does not contribute to appropriate regional development by evaluating and presenting the relationships between multiple regional indicators.

[0008] An object of the present invention is to provide a regional index evaluation device and a regional index evaluation method that enable balanced regional development and regional planning by evaluating the relationships among multiple regional indexes regarding the provision of facilities and services in a certain area. [Means for solving the problem]

[0009] The configuration of the regional index evaluation device of the present invention is preferably a regional index evaluation device that evaluates regional indexes when facilities are installed at locations in a region, and holds an index-facility relationship evaluation table that stores facility types related to regional indexes, evaluation values ​​of relationships between regional indexes, and evaluation values ​​of relationships between regional indexes and facility types, and a location information table that stores evaluation values ​​of relationships between regional locations and facility types, and creates an integrated graph with weights on each edge, with regional indexes, facility types, and locations as nodes, regional indexes and regional indexes as respective nodes, edges with regional indexes and facility types as respective nodes, and edges with facility types and locations as respective nodes, calculates feature vectors of each node in the integrated graph, calculates connection strength between nodes from the feature vectors of nodes corresponding to the regional indexes, and evaluates the strength of the relationship between the regional indexes based on the connection strength between the nodes. Effect of the Invention

[0010] According to the present invention, it is possible to provide a regional index evaluation device and a regional index evaluation method that enable balanced regional development and regional planning by evaluating the relationships among multiple regional indexes regarding the provision of facilities and services in a certain area. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 is a functional configuration diagram of a regional indicator evaluation system. [Diagram 2] FIG. 1 is a diagram showing the hardware and software configuration of a regional index evaluation device. [Diagram 3] FIG. 13 is a diagram showing an example of an index / facility related information table. [Figure 4] FIG. 4 is a diagram illustrating an example of a location information table. [Diagram 5] FIG. 13 is a diagram illustrating an example of an integrated graph information table. [Figure 6] FIG. 13 is a diagram illustrating an example of a feature vector table. [Figure 7] FIG. 13 is a diagram illustrating an example of an inter-node connection strength information table. [Figure 8] FIG. 1 illustrates an integrated graph. [Figure 9] 1 is a flowchart showing the processing of the regional index evaluation system. [Figure 10] FIG. 13 is a diagram showing an example of an index / facility information setting screen. [Figure 11] FIG. 13 is a diagram showing an example of a regional index evaluation result screen. [Figure 12] FIG. 13 is a diagram illustrating an example of a general index information table. [Figure 13] FIG. 13 illustrates an extended integration graph. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, embodiments of the present invention will be described with reference to FIGS. 1 to 13. FIG.

[0013] [Embodiment 1] Hereinafter, a first embodiment of the present invention will be described with reference to FIGS. First, the configuration of the regional index evaluation system of the first embodiment will be described with reference to FIG. 1 and FIG.

[0014] In this embodiment, an example will be described in which a regional index at a specific point in a region is evaluated by the regional index evaluation system shown in FIG. As shown in FIG. 1, the regional index evaluation system includes a regional index evaluation device 100, a user terminal 200, and a database server 300 connected via a network 5.

[0015] The regional index evaluation device 100 is a device for evaluating the relationships between regional indexes, and refers to an index / facility related information table 501 (details will be described later) obtained from evaluation materials and information on related facilities in the region held by the connected user terminal 200, and a location information table 502 (details will be described later) which stores evaluation information on facilities at each location held by the database server 300, and calculates the relationships and their strength between regional indexes, the relationships and their strength between regional indexes and locations, and the relationships and their strength between regional indexes and facilities.

[0016] The user terminal 200 is a terminal that stores the index / facility related information table 501 based on information input by a user to the regional index evaluation device 100, and also displays the results of evaluating the regional index to the user.

[0017] The database server 300 is a server that holds evaluation information for facilities existing at each location as a location information table 502.

[0018] The regional index evaluation device 100 has, as functional units, a user terminal I / F unit 111, a database I / F unit 112, a data linking unit 101, a graph reconstruction unit 102, a vector calculation unit 103, and a machine learning unit 104.

[0019] The user terminal I / F unit 111 is a functional unit for performing interface processing with the user terminal 200, passing data of the index / facility related information table 501 held by the user terminal 200 to the data linking unit 101, and passing results output by the graph reconstruction unit 102 and the vector calculation unit 103 to the data setting / result display unit 201 of the user terminal 200. The database I / F unit 112 is a functional unit for performing interface processing with the database server 300, passing data of the location information table 502 to the data linking unit 101. The data linking unit 101 is a functional unit for linking data of the index / facility related information table 501 held by the user terminal 200 with data of the location information table 502 held by the database server 300, and storing the data in an integrated graph information table 503 (described in detail later). The integrated graph is an undirected graph defined by the integrated graph information table, and is a graph that links, in layers, the relationships between the regional indexes and facilities stored in the information table index / facility related information table 501, their manually input evaluation values, and the information on the facilities and points held in the point information table 502. Details of the integrated graph will be described later.

[0020] The graph reconstructing unit 102 is a functional unit that reconstructs the graph based on the strength between vectors, the evaluation values ​​between regional indexes in the index and facility related information table 501, and the evaluation values ​​between regional indexes and facilities, and evaluates the relationship between feature vectors for the relationship between regional indexes that was not defined in the index and facility related information table 501, and also calculates the inter-node strength of the nodes between regional indexes that were defined in the index and facility related information table 501, thereby setting values ​​in the inter-index relationship graph information table as data for a new graph.

[0021] The vector calculation unit 103 is a functional unit that calculates the strength of a relationship for the vectors in the feature vector table 504 recorded as the learning results, using a calculation method such as vector distance.

[0022] The machine learning unit 104 is a functional unit that performs machine learning using the data in the integrated graph information table 503 as learning data, and stores the regional index, facility type, and location as feature quantities of nodes in a graph structure as data in the feature vector table 504 using, for example, a graph embedding method. (Note that the graph embedding method is described in, for example, Graph Embeddings: Primo Godec, Dec 31, 2018, Graph Embeddings - The Summary<https: / / qiita.com / kk31108424 / items / 3525e9ada907466902e6> See .)

[0023] The user terminal 200 has a data setting and result display unit 201. The data setting and result display unit 201 allows the user to input and retain the name of a regional index, the relationship between a certain regional index and other regional indexes, the facility type settings related to the regional index, and the evaluation values ​​related thereto, and sets the roughest relationship between the regional index and other regional indexes that the user can know, and the evaluation values ​​of the facility type settings related to the regional index, as basic information before the regional index evaluation device performs calculations, and stores them in the index and facility related information table 501. The data setting and result display unit 201 also outputs and displays the graphs calculated by the regional index evaluation device 100 and related result information.

[0024] Next, the hardware and software configuration of the regional index evaluation device will be explained using FIG. As shown in Figure 2, the regional index evaluation device 100 has a hardware configuration including a central processing unit 402, a main memory device 403, an internal bus 404, a bus I / F 405, an external bus 406, a screen output I / F 407, a user input I / F 408, a large-capacity memory device 409, a communication I / F 410, and an input / output I / F 411.

[0025] The central processing unit (CPU) 402 is a device that controls each part of the regional index evaluation device 100, loads programs installed in the auxiliary storage device 430 into the main storage device 403, and performs calculations for program execution.

[0026] The main memory 403 is realized by a semiconductor device such as RAM, and is used as a processing area during program execution and as a temporary storage area for data. For example, it temporarily stores basic programs such as an OS (Operating system), programs for implementing processing in each device, utility programs for implementing the functions of the regional index evaluation device 100 described below, and work data referenced by these programs.

[0027] Here, the central processing unit 402 and the main memory device 403 are connected by an internal bus 404 , and the internal bus 404 is connected to an external bus 406 via a bus I / F 405 .

[0028] The external bus 406 is connected to a screen output I / F 407, an input / output I / F 408, an auxiliary memory I / F 409, and a communication I / F 410, and data input / output between these I / Fs and the central processing unit 402 and main memory device 403 is performed via a bus I / F 405 and an internal bus 404.

[0029] The screen output I / F 407 is an I / F for connecting to a display 410, and outputs graphic data obtained via the external bus 406 to the display 410. The display 410 is, for example, an LCD (Liquid Crystal Display), and is a device capable of displaying an index / facility information setting screen, processing status, etc., which will be described later.

[0030] The input / output I / F 408 is an I / F for connecting to an input / output device, and performs input / output of data between the external bus 406 and the input / output device. The input / output device is a device for receiving input from a user, such as a keyboard 421 or a mouse 422, as shown in Fig. 2, or an input device for reading data from an external medium, such as a DVD drive or a card reader, or an output device, such as a printer, although not shown.

[0031] The auxiliary storage I / F 409 is an I / F for connecting to the auxiliary storage device 430 , and mediates data input / output from the auxiliary storage device 430 to the external bus 406 .

[0032] The auxiliary storage device 430 is, for example, a non-volatile storage device such as a large-capacity HDD (Hard Disk Drive) or SDD (Solid State Drive). A data linking program 231, a graph reconstruction program 232, a vector operation program 232, and a machine learning program 234, which are programs for realizing the functions of the basic program (not shown) and the regional index evaluation device 100, are installed in the auxiliary storage device 430. The data linking program 231, the graph reconstruction program 232, the vector operation program 232, and the machine learning program 234 are programs for realizing the functions of the data linking unit 101, the graph reconstruction unit 102, the vector operation unit 103, and the machine learning unit 104, respectively.

[0033] The auxiliary storage device 430 also holds an integrated graph information table 503 , a feature vector table 504 , and an inter-node connection strength information table 505 .

[0034] The communication I / F 410 is an I / F for connecting to a communication device 440, and mediates data input / output from the communication device to the external bus 406. The communication device is, for example, a device such as a network interface card that communicates with an external server device, terminal, etc. via a network according to the Ethernet (registered trademark) standard.

[0035] Next, the data structure used in the regional index evaluation system of this embodiment will be described with reference to FIGS.

[0036] The index / facility related information table 501 is a table that stores information about regional indexes, facilities, relationship information between regional indexes, relationship information between regional indexes and facilities, evaluation values ​​of the relationships between regional indexes, and evaluation values ​​of the relationships between regional indexes and facilities, and is held in the user terminal 200.

[0037] The index / facility relation information table 501 includes fields for a regional index ID 501a, a regional index name 501b, a related regional index ID 501c, a related facility type 501d, and a relation evaluation value 501e.

[0038] The area index ID 501a stores an ID that uniquely represents each area index. The area index name 501b stores a name of the area index that is easy for the user to identify. The related area index ID 501c stores an ID of the area index that the user recognizes as being related to the area index of the area index ID 501. The related facility type 501d stores the name and ID of the facility type that the user recognizes as being related to the area index of the area index ID 501. Here, the facility type represents the type of facility, such as "sports facility," "community center," and "market," as shown in FIG. 3. The relationship evaluation value 501e stores a value that evaluates the relationship between the area index of the area index ID 501, which is predefined or input by the user, and another area index, or a value that evaluates the relationship between the area index of the area index ID 501, which is predefined or input by the user, and the facility type.

[0039] By making such settings, information such as which regional indicators are related to which facilities and the general nature of the relationships between regional indicators can be used as material for analyzing the relationships between regional indicators. For example, as shown in FIG. 3, for the regional indicator ID "K001," "QoL (Quality of Life) KPI (Key Performance Indicator)" is set as the regional indicator name, and "K002" is set as the related regional indicator ID related to that regional indicator. In addition, "sports facilities: F011, community centers: F012, parks: F013" are set as the related facility type IDs of the facility types related to that regional indicator (note that in this table example, names and IDs are shown as values, but only one of them may be used). By making such settings, it is shown that the regional indicator with the regional indicator ID "K001" and the name "QoLKPI" is related to another regional indicator ID "K002" (regional indicator name "Environmental KPI"), and the types of facilities related to that regional indicator are "sports facilities," "community centers," and "parks."

[0040] For the regional index with regional index ID "K002", the value of the relationship evaluation value 501e is "K001 / 30", indicating that the relationship evaluation value between the regional index with regional index ID "K002" and the regional index with regional index ID "K001" is "30." Furthermore, for the connection ID "K003", the value of the relationship evaluation value 501e is "F031 / 20", indicating that the relationship evaluation value between the regional index with regional index ID "K003" and the facility type "independent store" is "20."

[0041] This allows the user to set the strength of the relationship between regional indicators as an evaluation value, and in particular the strength of the connection between the regional indicator and the type of connected facility as an evaluation value.

[0042] The point information table 502 is a table that stores information on each point in a region, and is held in the database server 300.

[0043] As shown in FIG. 4, the point information table 502 includes fields for a point ID 502a, an adjacent point ID 502b, a facility type ID 502c, and an evaluation value 502d.

[0044] The location ID 502a stores an ID that identifies a location assigned to each block, such as X-chome, X-banchi. The adjacent location ID 502b stores, in list form, the IDs of locations geographically adjacent to the location indicated by the location ID 502a. The facility type ID 502c stores the name and ID of the facility type assumed to be placed at the location indicated by the location ID 502a. The evaluation value 502d stores an evaluation value when a facility of the facility type indicated by the facility type ID 502c is placed at the location indicated by the location ID 502a.

[0045] By preparing such table data, a graph structure can be constructed based on the adjacency relationships between locations and the connections between each location and facility information, and the evaluation of each facility can be expressed as the strength and weighting of the connection between the location and facility.

[0046] For example, the point with point ID "P001" is adjacent to the points with adjacent point IDs "P010, P011, P012, P013", and when a facility of the facility type "Community Center: F012" is placed, the evaluation value is "10".

[0047] In this example, a community center is located at a point on a certain point (X-chome, No. X), and the evaluation value when evaluated here (evaluation value under the local indicator name "QoL KPI" in Figure 3) is 10, which means, for example, that the number of people using the community center is 100 [people / month]. Also, when a facility of facility type "Market: F032" is placed at point ID "P010" adjacent to point ID "P001", the evaluation value (evaluation value under the local indicator name "Commerce KPI" in Figure 3) is 40, which means, for example, that when a market is placed at point ID "P010", city tax revenue will increase by 4 million yen / year.

[0048] The integrated graph information table 503 is a table for storing information on the integrated graph. The integrated graph is an undirected graph in which nodes are area indices, facility types, and locations, and edges are connections between area indices, area indices and facility types, and locations and facility types. A specific example of the integrated graph will be described later.

[0049] As shown in FIG. 5, the integrated graph information table 503 includes fields of a connection source ID 503a, a connection destination ID 503b, and a weight 503c.

[0050] A region index ID, a facility type ID, and a location ID are set on one side of the connection source ID 503a and the connection destination ID 503b, and define the nodes at both ends of the edge in a pair and the edge connecting them. Possible combinations of the values ​​of the connection source ID 503a and the connection destination ID 503b are (region index ID, region index ID), (region index ID, facility type ID), and (location ID, facility type ID) when listed with the order of the values ​​being the same. The weight 503c stores the value of the weight associated with the edge defined by the connection source ID 503a and the connection destination ID 503b. The weight of the edge corresponding to (region index ID, region index ID) and (region index ID, facility type ID) is set by referring to the value of the relationship evaluation value 501e in the index-facility related information table 501 in FIG. 3, and the weight of the edge corresponding to (region index ID, facility type ID) is set by referring to the value of the evaluation value 502d in the location information table in FIG. 4. In the example of FIG. 5, each evaluation value is used as is, but a function may be used that uses the evaluation value as a parameter and converts the edge weight to a larger value as the evaluation value increases, or the value of weight 503c may be set by referring to a table that defines the edge weight in correspondence with the evaluation value.

[0051] The feature vector table 504 is a table for storing information on feature vectors for nodes of an integrated graph, and has fields for a node ID 504a and a feature vector 504b.

[0052] The node ID 504a stores the value of one of the area index ID, facility type ID, and location ID of each node constituting the integrated graph. The feature vector 504b stores the feature vector corresponding to the node of the value of the node ID 504a. The feature vector corresponding to the node ID 504a can be obtained by performing machine learning on the graph represented by the data of the integrated graph information table 503. For example, a method such as DeepWalk, node2vec, or SDNE can be used as a node embedding method in the graph embedding method.

[0053] The inter-node connection strength information table 505 is a table that holds the inter-node connection strengths of the nodes of the integrated graph defined by the integrated graph information table 503 (obtained by evaluating the feature vectors at the nodes of the integrated graph; details will be described later), and as shown in FIG. 7, consists of fields for source ID 505a, destination ID 505b, inter-node connection strength 505c, and new strength setting flag 505d.

[0054] The connection source ID 505a and the connection destination ID 505b are stored by copying the values ​​of the connection source ID 503a and the connection destination ID 503b in the integrated graph information table 503. The inter-node connection strength 505c stores the inter-node connection strength of the node having the regional index ID of the connection source ID 505a and the connection destination ID 505b. The new strength setting flag 505d is set to 1 when the value of the weight 503c of the record corresponding to the connection source ID 503a and the connection destination ID 503b from which the connection source ID 505a and the connection destination ID 505b are copied is not defined, and is set to 0 otherwise.

[0055] Next, the integrated graph will be described with reference to FIG. As shown in Fig. 8, the integrated graph 600 has nodes having area index IDs (hereinafter simply "area index nodes"), nodes having facility type IDs (hereinafter simply "facility type nodes"), and nodes having location IDs (hereinafter simply "location nodes"), and is made up of edges connecting area index nodes with area index nodes, edges connecting area index nodes with facility type nodes, and edges connecting facility type nodes with locations. The weights set in Fig. 5 are defined for each edge.

[0056] Nodes and edges consisting of area indicator nodes can be classified into layers: area indicator layer 610; nodes consisting of facility type nodes can be classified into facility type layer 620; and nodes consisting of points can be classified into map layer 630.

[0057] Next, the processing of the regional index evaluation system will be described with reference to FIG. FIG. 9 is a flowchart showing the processing of the regional index evaluation system.

[0058] First, the user terminal 200 accepts input from the user on an index / facility information setting screen (details of which will be described later) of names for regional indexes, associated facility types, definitions of other related regional indexes, evaluation values ​​of the relationships between regional indexes, and evaluation values ​​of the relationships between regional indexes and facility types, and sets values ​​in the index / facility related information table 501 shown in Figure 3 (S101).

[0059] Next, the data linking unit 101 of the regional index evaluation device 100 reads data from the index-facility related information table 501 of the user terminal 200, reads related information between regional indexes from the regional index ID 501a and related regional index ID 501c, and related information between regional indexes and facility types from the regional index ID 501a and related facility types 501d, and reads information related to those evaluation values ​​from the relationship evaluation value 501e. The data linking unit 101 of the regional index evaluation device 100 also reads data from the location information table 502 held by the database server 300, reads information related to the location and facility type from the location ID 502a and facility type ID 502c, and reads the related evaluation value from the evaluation value 502d.

[0060] Then, based on the information read from the index-facility related information table 501 and the location information table 502, the data linking unit 101 of the regional index evaluation device 100 sets values ​​for the connection source ID 503a, connection destination ID 503b, and weight 503c in the integrated graph information table 503 shown in FIG. 5 to represent edges whose both ends are a regional index node-regional index node, edges whose both ends are a regional index node-facility type node, and edges whose both ends are a facility type node-location node, and their weights (S102).

[0061] Next, the machine learning unit 104 of the regional index evaluation device 100 uses the data in the set integrated graph information table 503 to calculate feature vectors of nodes for the integrated graph information using a machine learning method such as the graph embedding method described above, and stores the feature vectors for each node of the integrated graph information in the feature vector table 504 shown in Figure 6 (S103).

[0062] Next, the vector calculation unit 103 of the regional index evaluation device 100 refers to the feature vector for each node in the feature vector table 504, calculates the inter-node connection strength for the nodes of the integrated graph (S104), and sets the value in the inter-node connection strength information table 505 shown in Figure 7.

[0063] The inter-node connection strength can be calculated, for example, from the Euclidean distance of the feature vector for each node shown in the following (Equation 1).

[0064]

number

[0065] The inter-node connection strength of the nodes corresponding to each feature vector is set so that the smaller the Euclidean distance of the feature vector for each node, the greater the inter-node connection strength. In order to obtain the inter-node connection strength from the Euclidean distance of the feature vector for each node, a conversion function may be used, or the inter-node connection strength may be obtained by referring to a table that defines the inter-node connection strength in correspondence with the Euclidean distance of the feature vector.

[0066] Next, the graph reconstruction unit 102 of the regional index evaluation device 100 sets a value in the inter-node connection strength information table 505 based on the inter-node connection strength for the nodes of the integrated graph calculated in S104 (S105). At that time, if the evaluation value between regional indexes is not set in the index / facility related information table 501 in Fig. 3 (the corresponding weight is not defined in the integrated graph information table 503), 1 is set.

[0067] Next, the user terminal 200 displays the results of the regional index evaluation, in particular the inter-index relationship graph, on the regional index evaluation result screen (details of which will be described later) based on data from the index / facility related information table 501, location information table 502, integrated graph information table 503, inter-node connection strength information table 505, and regional map information (not shown) (S106).

[0068] Next, the user interface provided by the regional index evaluation device will be described with reference to FIG. 10 and FIG.

[0069] The index / facility information setting screen 800 is a screen for inputting information related to regional indexes and facilities from the user terminal 20, and the input values ​​are reflected as data in the index / facility related information table 501. The index / facility information setting screen 800 includes, for example, an index / facility information setting area 810, an integrated graph display area 820, and an evaluation execution button 830, as shown in FIG.

[0070] The index / facility information setting area 810 is a field for inputting information regarding regional indicators and facility types, and consists of a regional indicator ID input field 811, a regional indicator name input field 812, a related regional indicator ID input field 813, a related facility type input field 814A, a related facility type ID input field 814B, a related ID input field 815A, an evaluation value input field 815B, and a set / update button 816.

[0071] The regional index ID input field 811, the regional index name input field 812, the related regional index ID input field 813, (related facility type name input field 814A, related facility type ID input field 814B), (related ID input field 815A, evaluation value input field 815B) are input fields corresponding to the fields of the regional index ID 501a, the regional index name 501b, the related regional index ID 501c, the related facility type 501d, and the relationship evaluation value 501e, respectively, of the index / facility related information table 501.

[0072] After the user inputs these values, the user can press a set / update button 816 with a pointing device such as a mouse to set new values ​​or update the values.

[0073] The integrated graph display area 820 is an area for displaying the integrated graph 600 created based on the index and facility information input information.

[0074] As shown in FIG. 10, the integrated graph display 821 displays a region index layer 821A, a facility type layer 821B, and a map layer 821C in separate layers.

[0075] The area indicator layer 821A displays area indicator nodes and the connections of the associated nodes to each other.

[0076] The facility type layer 821B displays facility type nodes, and the integrated graph display 821 displays connections between the facility type nodes of the integrated facility type layer 821B and the area indicator nodes of the upper layer, area indicator layer 821A.

[0077] The map layer 821C displays location nodes of the facility locations on the map. The integrated graph display 821 also displays connections between the location nodes of the map layer 821C and the upper layer, the integrated facility type layer 821B.

[0078] In the integrated graph display area 820, to facilitate editing, the user may click with a pointing device such as a mouse to indicate a node on each integrated graph, whereupon the setting contents are output to the index / facility information setting area 810 and can be modified as necessary.

[0079] The setting / update button 830 is a button that the user presses after completing the settings in the index / facility information setting area 810 and checking the integrated graph displayed in the integrated graph display area 820 to start evaluation using the regional index evaluation device 100.

[0080] The regional index evaluation result screen 900 is a screen for displaying the evaluation results of the regional index of the regional index evaluation device 100, and includes, for example, a current status evaluation designation area 910, a facility change evaluation designation area 920, an evaluation display area 930, and a regional index relationship strength display area 940.

[0081] The current status evaluation specification area 910 is an input area for specifying a location ID and displaying each regional index in question. The user inputs a location ID into the location ID input field 911 and presses the result display button 912 with a pointing device such as a mouse, thereby causing the score of the corresponding regional index to be displayed in the evaluation display area 930.

[0082] The facility change evaluation designation area 920 is an area for displaying the changes in each regional index when the current facility type of the location specified by the location ID is changed to another facility type. The user enters a location ID in the location ID input field 921 and a facility type ID in the facility type input field 922, and presses the result display button 923 with a pointing device such as a mouse, whereby the score of the corresponding regional index is displayed in the evaluation display area 930.

[0083] The evaluation display area 930 is an area that displays the results of evaluation of related regional indicators when a certain facility type is placed at a certain point, and for the point on the map specified in the current situation evaluation specification area 910 or the facility change evaluation specification area 920, the scores of the related regional indicators are displayed as in a speech bubble 931 saying "The current situation evaluation is..." and a speech bubble 932 saying "If a facility type community center is placed...", as shown in Fig. 11. Note that in order to specify the evaluation of the regional indicators in the speech bubble like this, a method of specifying the point on the map by clicking with a mouse or the like may be used.

[0084] The evaluation corresponding to this point ID can be found from the value of the node connection strength 505c of the area index ID-facility type ID in the node connection strength graph information table 505 and the value of the node connection strength 505c of the facility type ID-area index ID.

[0085] When the facility change evaluation designation area 920 is designated, the score of the regional index in the speech bubble 932 may be displayed as is, or, as shown in FIG. 11, an UP or DOWN value may be displayed as the difference with the speech bubble 931.

[0086] The regional index relationship strength display area 940 is an area that displays the strength of the relationships between regional indexes before and after relationship analysis, comparing them, and also displays previously unseen hidden relationships and the strength of those relationships between the regional indexes as results.

[0087] For example, the pre-relationship analysis display area 941 displays the evaluation value 501 of the regional index from the index / facility related information table 501 as the value of the relationship strength of the regional index, and the post-relationship analysis display area 942 displays the value of the inter-node bond strength 505c between each regional index node in the inter-node bond strength information table 505 as the value of the relationship strength of the regional index.

[0088] For example, before the relationship analysis, the relationship strength between the QoL KPI and the environmental KPI was 30, and the relationship strength between the environmental KPI and the commercial KPI was 20. For the relationship between the commercial KPI and the QoL KPI, only a connection relationship was defined, and the value of the relationship strength (evaluation value in FIG. 3) was not set. In contrast, after the relationship analysis, the relationship strength between the QoL KPI and the environmental KPI and the commercial KPI is updated to 60, and the relationship strength between the environmental KPI and the commercial KPI is updated to 10. For the relationship between the commercial KPI and the QoL KPI, the relationship strength between the commercial KPI and the QoL KPI is newly displayed as a hidden connection relationship that was not visible before the analysis, with the relationship strength between the commercial KPI and the QoL KPI being 30. Such regional index relationships are identified by looking at the new strength setting flag 505d in the inter-node connection strength information table 505.

[0089] As described above, the regional index evaluation device of this embodiment makes it possible to evaluate not only facilities in a region, but also services (mobile sales and events) that are provided for a limited time in a certain location, while taking into consideration the interrelationships between multiple regional indexes according to the facility type of each service. Using this result, in public policies and private business policies in the region, it is possible to carry out layout plans according to the facility type at each location while balancing each regional index.

[0090] [Embodiment 2] A second embodiment of the present invention will be described below with reference to FIGS.

[0091] In the first embodiment, a regional index evaluation device that evaluates the relationship between indexes when arranging facilities of a certain facility type at each point in a region has been described.

[0092] In this embodiment, an example of a regional index evaluation device that analyzes regional indexes not only for a single region (e.g., an administrative city) but also for multiple regions (an area that combines multiple neighboring cities) that bundle the region will be described. The configuration and concept of the regional index evaluation device of this embodiment are almost the same as those of the first embodiment, so the differences will be mainly described.

[0093] In this embodiment, a different level of index is provided from the regional indexes defined in the index / facility related information table 501 in Fig. 3 of embodiment 1. For example, in Fig. 3, in addition to the regional indexes named QoL KPI, environment KPI, and commerce KPI, indexes named general QoL KPI, general environment KPI, and general commerce KPI are provided as general indexes one level above them, and the QoL KPI, environment KPI, and commerce KPI are used as indexes for evaluating the installation of facilities in a certain region, and the general indexes are used for evaluation of a layer that combines multiple regions.

[0094] For this purpose, a general index information table 510 having a data structure shown in FIG. 12 holds information on general indexes and relationships between general indexes and regional indexes.

[0095] The general index information table 510 is a table that stores information about general indexes and the relationships between general indexes and regional indexes, and as shown in FIG. 12, it consists of fields for general index ID 510a, general index name 510b, related regional index ID 510c, and relationship evaluation value 510e.

[0096] The general index ID 510a stores an ID that uniquely represents each general index. The general index name 510b stores a name of the general index that is easy for the user to identify. The related index ID 510c stores a general index ID or a regional index ID that the user recognizes as being related to the index with the ID of the general index ID 510. The relationship evaluation value 510e stores a value that evaluates the relationship between the index with the ID of the general index ID 510 and the regional index, which is predefined or input by the user.

[0097] Next, the concept of the integrated graph of this embodiment will be described with reference to FIG. The integrated graph of this embodiment is an expanded version of the integrated graph of FIG. 8 of the first embodiment (hereinafter, referred to as the "extended integrated graph").

[0098] The extended integrated graph 700 has a structure shown in Fig. 13, and a general indicator layer 710 is provided as an upper layer to the index-related layer 610 of the integrated graph 700 in Fig. 8. The nodes of the general indicator layer (hereinafter simply "general indicator nodes") are connected to each other. Here, the layer corresponding to the index-related layer 610 in the first embodiment is the area indicator layer 720, and the area indicator nodes of this layer (hereinafter simply "area indicator nodes") can be connected to the general indicator nodes of the general indicator layer 710.

[0099] Also, there may be nodes expressing region-specific regional indicators, such as "X City Tourism KPI" shown in FIG.

[0100] The edges between the nodes have their own weights and are represented by a data structure such as integrated graph information table 503 in FIG. 5, as in the first embodiment.

[0101] In this embodiment, the evaluation of the strength of the relationship between the general indexes expanded and the evaluation of the strength of the relationship between the general index and the regional index can be obtained in the same manner as in embodiment 1. In this embodiment, an expanded integrated graph can be displayed on the index / facility information setting screen 800 shown in Fig. 10.

[0102] In this embodiment, by introducing a general index that is higher than the regional index used when facilities of a facility type are installed in a region, it becomes possible to analyze the relationship between regional indexes not only for a single region, but also for a broader region that encompasses that region. [Explanation of symbols]

[0103] 5…Network, 100: Regional index evaluation device, 101: Data linking unit, 102: Graph reconstruction unit, 103... vector calculation unit, 104... machine learning unit, 111... user terminal I / F unit, 112... database I / F unit, 200...user terminal, 201...data setting and result display unit, 300: Database server, 501: Index / facility related information table; 502: Location information table; 503: Integrated graph information table; 504: Feature vector table; 505: Inter-node connection strength information table; 510...General indicator information table, 600...integrated graph, 610...regional indicator layer, 620...facility type layer, 630...map layer, 700…Extended Integrated Graph, 710: General indicator layer, 720: Regional indicator layer

Claims

1. A regional index evaluation device for evaluating a regional index when a facility is installed at a location in a region, comprising: an index / facility association evaluation table that stores the regional index and the associated facility type, the evaluation value of the relationship between the regional indexes, and the evaluation value of the relationship between the regional index and the facility type; A location information table is stored to store evaluation values ​​of the relationships between the locations of the area and the facility types; based on information in the index / facility related evaluation table and the location information table, an integrated graph is created in which the regional index, the facility type, and the location are nodes, the regional index and the regional index are each a node, edges are the regional index and the facility type are each a node, and edges are the facility type and the location are each a node, and each edge has a weight; Calculating a feature vector for each node of the integrated graph; A regional index evaluation device characterized by calculating the connection strength between nodes from the feature vectors of the nodes corresponding to the regional indexes, and evaluating the strength of the relationship between the regional indexes based on the connection strength between the nodes.

2. The regional index evaluation device according to claim 1, characterized in that the evaluation value of the relationship between regional indexes stored in the index / facility relation evaluation table is compared with the value of the strength of the relationship between the regional indexes evaluated based on the connection strength between the nodes, and the comparison is displayed on a display means.

3. 2. The regional index evaluation device according to claim 1, wherein the structure of the integrated graph is displayed by forming the regional index nodes as a regional index layer, the facility type nodes as a facility type layer, and the location nodes as a map layer.

4. Furthermore, a general index table is held, which stores evaluation values ​​of general indexes other than the indexes defined in the index / facility related evaluation table and the regional indexes defined in the index / facility related evaluation table; 2. The regional index evaluation device according to claim 1, wherein the integrated graph further includes a node of the general index and includes edges connecting the general index and the general index as each node.

5. The regional index evaluation device according to claim 4, characterized in that the structure of the integrated graph is displayed by making the nodes of the general index into a general index related layer, the nodes of the regional index into a regional index layer, the nodes of the facility types into a facility type layer, and the nodes of the locations into a map layer.

6. 2. The regional index evaluation device according to claim 1, wherein the feature vector of each node of the integrated graph is calculated by machine learning.

7. 7. The regional index evaluation device according to claim 6, wherein a feature vector of each node of the integrated graph is calculated by a graph embedding method.

8. A regional index evaluation device that evaluates regional indexes when a facility is installed at a location in a region, The regional index evaluation device includes: an index / facility association evaluation table that stores the regional index and the associated facility type, the evaluation value of the relationship between the regional indexes, and the evaluation value of the relationship between the regional index and the facility type; A location information table is stored to store evaluation values ​​of the relationships between the locations of the area and the facility types; a step in which the regional index evaluation device creates an integrated graph based on information in the index / facility relation evaluation table and the location information table, with the regional index, the facility type, and the location as nodes, the regional index and the regional index as respective nodes as edges, the regional index and the facility type as respective nodes as edges, and the facility type and the location as respective nodes as edges, with weights on each edge; calculating a feature vector for each node of the integrated graph; a step of calculating a connection strength between nodes from the feature vector of the node corresponding to the regional index, and evaluating the strength of the relationship between the regional indexes based on the connection strength between the nodes.

Citation Information

Patent Citations

  • District evaluation index processing system, district evaluation index processing program and district evaluation index processing method

    JP2016206772A