Updating device, updating method, and updating program

Abstract

To appropriately update a vector indicating a feature.SOLUTION: An updating device according to the present application comprises: a coupling unit that couples between a user and an area visited by the user on the basis of a visit history of the user in a bipartite graph including a set of users and a set of areas; and an updating unit that updates a user vector or an area vector on the basis of the user vector indicating a feature of the user based on an action of the user on a network and the area vector indicating a feature of the area coupled to the user by the coupling unit.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an update device, an update method, and an update program.

Background Art

[0002] Conventionally, for example, a learning device that learns a model for generating a vector indicating the characteristics of a user from the user's behavior history on the Internet has been proposed (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the prior art, for example, there is room for improvement in updating the vector. Specifically, since the vector is generated or updated based on the behavior on the Internet, the update of the vector for the behavior in the real world has not been considered.

[0005] The present invention has been made in view of the above, and an object thereof is to provide an update device, an update method, and an update program capable of appropriately updating a vector indicating characteristics.

Means for Solving the Problems

[0006] In order to solve the above-described problems and achieve the object, an update device according to the present invention includes a combining unit that combines a user and a region visited by the user based on the access history of the user in a bipartite graph including a set of users and a set of regions, a user vector indicating characteristics of the user based on the user's actions on the network, and an update unit that updates the user vector or the region vector based on the region vector indicating characteristics of the region combined with the user by the combining unit.

Advantages of the Invention

[0007] According to the present invention, a vector indicating characteristics can be appropriately updated.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Modes for Carrying Out the Invention

[0009] Hereinafter, embodiments for implementing the update device, update method, and update program according to the present application (hereinafter referred to as "embodiments") will be described in detail with reference to the drawings. Note that the update device, update method, and update program according to the present application are not limited by these embodiments.

[0010] [Embodiments] [1. Information Processing] First, an example of information processing according to the embodiment will be described with reference to FIG. 1. FIG. 1 is a diagram showing an example of information processing according to the embodiment.

[0011] As shown in FIG. 1, the update device 10 according to the embodiment acquires the behavior information of a user (not shown) on the network and the behavior information in the real world, and outputs a user vector or a region vector.

[0012] Here, the behavior information on the network is, for example, information regarding the usage history of various portal sites, and includes information regarding purchase history, search history, browsing history, etc. Also, the behavior information in the real world is, for example, information regarding the visit history of each user in the real world, and includes information regarding position information obtained by the user's user terminal (for example, a smartphone).

[0013] The user vector is a vector indicating the characteristics of the user based on the user's behavior on the network. For example, the user vector is generated by a user vector conversion model that is learned to generate similar user vectors as the behavior information on the network becomes more similar. Note that vectors being similar means that the cosine distance of the vectors is similar.

[0014] A region vector is a vector that represents the characteristics of a region in the real world. For example, the region vector is generated by a region vector conversion model that is trained to output more similar vectors as the proximity of the region itself, the similarity of the facilities located therein, the similarity of the roads, and the similarity of the user attributes of the visiting users increases. Note that the region is, for example, a divided region obtained by dividing map information along a predetermined rule (e.g., in a mesh shape). Therefore, in the bipartite graph described later, regions will also be connected according to their geographical positions.

[0015] First, the update device 10 according to the embodiment acquires the action information of each user on the network and the action information in the real world (steps S01, S02). Subsequently, the update device 10 connects the nodes in a bipartite graph including a set of users and a set of regions (step S03).

[0016] Here, the bipartite graph is a graph in which each user and each region are nodes. For example, user vectors Vu1 to Vu3 of the corresponding user are stored in the node corresponding to the user, and region vectors Va1 to Va3 of the corresponding region are stored in the node corresponding to the region.

[0017] The update device 10 connects the node corresponding to the user and the node corresponding to the region visited by the user by an edge based on the visit history (action information in the real world) of the user. Also, in the bipartite graph, regions are connected according to their geographical positions. Thereby, the update device 10 can construct a network that links the characteristics of the action information of the user on the network and the characteristics of the regions in the real world visited by the user.

[0018] Subsequently, the update device 10 updates the vectors (user vector and area vector) based on the combined user vector and area vector (step S04). For example, the update device 10 updates the user vector and the area vector by means of GCNs (Graph Convolutional Networks) for a bipartite graph. GCNs is a method of performing convolution on graph data. Convolution by graph data adds the weighted features of adjacent nodes to the features of a single node in the graph.

[0019] That is, the features of adjacent area vectors are added to the user vector, and the features of adjacent user vectors and other adjacent area vectors are added to the area vector. Therefore, the user vector and the area vector adjacent to the user vector are updated so that both vectors are similar to each other.

[0020] In the example shown in the figure, for example, the case where the user vector Vu1 and the area vector Va1 are updated with the weight w1, the user vector Vu1 and the area vector Va2 are updated with the weight w2, and the area vector Va1 and the area vector Va2 are updated with the weight w3 is shown.

[0021] The weights w1 and w2 between the user vector Vu1 and the area vectors Va1 and Va2 are values set based on, for example, the stay frequency and stay time of the corresponding user. That is, when the user has a higher stay frequency or a longer stay time in area B than in area A, the weight of the user vector corresponding to the user for the area vector corresponding to area B is stronger than the weight for the area vector corresponding to area A. The weight w3 between the area vector Va1 and the area vector Va2 is a value set according to, for example, the distance and positional relationship between the corresponding areas. For example, the closer the areas are to each other, the stronger the weight between them.

[0022] For example, when the update device 10 updates the user vector and the area vector by means of GCNs, it outputs the updated user vector or area vector (step S05). For example, the user vector output from the update device 10 is utilized for various recommendations on the network and the like, and the user vector output from the update device 10 is utilized for, for example, land development and the like.

[0023] As described above, the update device 10 according to the embodiment generates a bipartite graph connecting the user vector and the area vector by using the access history of the user, and updates the user vector and the area vector by means of GCNs for the bipartite graph.

[0024] Therefore, in the update device 10 according to the embodiment, the area vector indicating the characteristics of the real-world area can be reflected in the user vector indicating the characteristics of the actions on the network, and the user vector can be reflected in the area vector.

[0025] Therefore, according to the update device 10 according to the embodiment, the vector indicating the characteristics can be appropriately updated.

[0026] 〔2. Configuration Example of Update Device〕 Next, a configuration example of the update device 10 will be described with reference to FIG. 2. FIG. 2 is a block diagram showing a configuration example of the update device 10 according to the embodiment. As shown in FIG. 2, the update device 10 includes a communication unit 110, a control unit 120, and a storage unit 130.

[0027] The communication unit 110 is realized, for example, by a NIC (Network Interface Card) or the like. Then, the communication unit 110 transmits and receives information to and from an external device via a network such as various wireless communication networks such as 4G (Generation), 5G, LTE (Long Term Evolution), Wifi (registered trademark) or wireless LAN (Local Area Network), or various wired communication networks.

[0028] The storage unit 130 is implemented by, for example, a semiconductor memory element such as a RAM or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 130 also includes a user vector model storage unit 131, a user vector storage unit 132, a region vector model storage unit 133, a region vector storage unit 134, a map information storage unit 135, and a bipartite graph storage unit 136.

[0029] The user vector model storage unit 131 stores a user vector model. The user vector model is a learning model that converts information about a user into a user vector. For example, when the update device 10 inputs information about the behavior history of each user on the network into the model, the model is trained to output a user vector that is similar to the user vectors of other users whose information is similar.

[0030] Here, the information about the behavior history on the network includes, for example, information related to purchase history, search history, browsing history, similarity of the objects of purchase, search, or browsing, the time (or period) of purchase, search, or browsing, etc. on a portal site. Also, the model may be a model that generates a user vector according to information about the user attributes in addition to the information about the usage history of each user on the network.

[0031] Note that the user attributes here may be so-called demographic attributes such as age, gender, residential area, education level, family composition, workplace, annual income, etc., or so-called psychographic attributes such as hobbies, preferences, lifestyle, etc.

[0032] The user vector storage unit 132 stores user vectors. FIG. 3 is a diagram showing an example of the information stored in the user vector storage unit 132 according to the embodiment. As shown in FIG. 3, the user vector storage unit 132 includes information such as "user ID" and "user vector".

[0033] In the "User ID" field, identification information for identifying the user is stored. In the "User Vector" field, the user vector corresponding to the user identified by the user ID is stored. In FIG. 3, an example is shown where the user vector is two-dimensional vector data "10, 24", but the user vector may be multi-dimensional (N-dimensional) vector data and is not limited to two dimensions.

[0034] The area vector model storage unit 133 stores the area vector model. The area vector model is a learning model that converts information about the area into an area vector. For example, when the update device 10 inputs information about the area into the model, the model is trained to output a user vector that is similar to the user vectors of other users whose information is similar.

[0035] Note that the information about the area here includes information about the location facility, road information, section information, etc. Also, the area vector model may be a model that generates an area vector by averaging the user vectors of each visited user.

[0036] The area vector storage unit 134 stores the area vector. FIG. 4 is a diagram showing an example of the area vector storage unit 134 according to the embodiment. The area vector storage unit 134 includes information such as "Area ID" and "Area Vector".

[0037] In the "Area ID" field, identification information for identifying the area is stored. In the "Area Vector" field, the area vector of the user identified by the area ID is stored. In FIG. 4, an example is shown where the area vector is two-dimensional vector data "33, 15", but the user vector may be multi-dimensional (N-dimensional) vector data and is not limited to two dimensions.

[0038] The map information storage unit 135 stores map information indicating the positional relationship between each area. For example, based on such map information, the weights of the area vectors between areas will be set.

[0039] The bipartite graph storage unit 136 stores the bipartite graph illustrated in FIG. 1. As shown in FIG. 1, the bipartite graph is a graph including a set of users and a set of regions. In the bipartite graph, each user and each region become nodes, respectively, and the edges connecting users and regions are set based on the access history, and the edges connecting regions are set based on the geographical relationship between regions.

[0040] The control unit 120 is, for example, a controller, and is realized by various programs stored in the storage device inside the updating device 10 being executed with the RAM as a working area by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like. Further, the control unit 120 is a controller and is realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0041] As shown in FIG. 2, the control unit 120 includes a user vector generation unit 121, a region vector generation unit 122, a combination unit 123, a setting unit 124, an update unit 125, and a provision unit 126.

[0042] The user vector generation unit 121 generates a user vector indicating the characteristics of a user based on the actions of the user on the network. For example, the user vector generation unit 121 acquires information regarding the actions of the user on the network from a portal site or a user terminal held by each user at a predetermined cycle.

[0043] The user vector generation unit 121 generates a user vector by inputting the acquired information regarding the actions of the user on the network into the user vector model stored in the user vector model storage unit 131. Further, the user vector generation unit 121 stores the generated user vector in the user vector storage unit 132.

[0044] The area vector generation unit 122 generates an area vector based on information about the facilities, roads, and sections of the area. For example, the area vector generation unit 122 obtains the current map information of each area from an external device such as a server, and inputs the obtained map information into the area vector model stored in the area vector model storage unit 133 to generate an area vector.

[0045] Note that the area vector generation unit 122 may generate an area vector based on the user vector of the user who visited the area. In this case, for example, the area vector generation unit 122 may, for each area, weight the user vectors of each user based on the visit frequency, visit time, etc. of the user, and then generate the averaged value as the area vector. Also, the area vector generation unit 122 may generate one area vector for each area using the above map information and the above user vector.

[0046] The coupling unit 123 couples the node corresponding to the user and the node corresponding to the area in the bipartite graph based on the visit history of the user. The coupling unit 123 obtains, for example, the position information (e.g., GPS positioning information) of each user terminal at a predetermined period, and calculates the visit history of each user to each area.

[0047] Subsequently, the coupling unit 123 couples the node corresponding to the user and the node corresponding to the area visited by the user based on the calculated visit history. Also, the coupling unit 123 couples the areas according to the geographical relationship between the areas. That is, the coupling unit 123 couples the areas by coupling each area with its adjacent areas.

[0048] The setting unit 124 sets a weight related to the update between the user and the area according to the relationship between the user and the area based on the access history. Here, the weight related to the update is the degree of influence that a user vector (or area vector) exerts on an adjacent area vector (or user vector). When updating the vector, the stronger the weight related to the update, the stronger the influence of both vectors on each other's vectors.

[0049] For example, the setting unit 124 learns the relationship between the user and the area based on the access history of each user, and sets a weight related to the update between the user and the area according to the learned relationship.

[0050] The relationship between the user and the area here indicates the connection between the user and the area in the real world. For example, the higher the frequency of the user's visit to a specific area or the longer the visit time, the stronger the connection between the two.

[0051] Therefore, the setting unit 124 analyzes the access history using a predetermined algorithm and sets a weight related to the update between the user and the area.

[0052] Also, for example, the setting unit 124 may be set so that the weight related to the update between the user and the area gradually weakens over time. That is, in this case, the newer the date and time of the visit, the stronger the weight, and as the date and time of the visit gets older, the weight is set to weaken.

[0053] Also, the combining unit 123 may use either a valid edge or an invalid edge for the edge that connects the node corresponding to the user and the node corresponding to the area. For example, in the case of a valid edge, it is possible to design a bipartite graph in which the user vector influences the area vector, but the area vector does not influence the user vector.

[0054] The update unit 125 updates the user vector or the area vector based on the user vector indicating the characteristics of the user based on the user's actions on the network and the area vector indicating the characteristics of the area coupled to the user by the coupling unit 123.

[0055] For example, the update unit 125 updates each user vector and each area vector by means of GCNs for the bipartite graph stored in the bipartite graph storage unit 136. At this time, the update unit 125 may update only one of the user vector and the area vector.

[0056] That is, for example, the update unit 125 may update the area vector based on the user vector, or may update the user vector based on the area vector. Also, the update timing and update period of the user vector may be set separately from the update timing and update period of the area vector.

[0057] In addition, the update unit 125 updates each vector according to the weight related to the update set by the setting unit 124. The stronger the weight related to the update, the greater the mutual influence of the two vectors, and the weaker the weight related to the update, the more difficult it is for the two vectors to influence each other.

[0058] Here, an example of the update process will be described with reference to FIG. 5. FIG. 5 is a schematic diagram of the user vector and the area vector according to the embodiment. In the example shown in FIG. 5(a), the user vector Vu and the area vector Va are vectors with the same dimension, and the update unit 125 multiplies the user vector Vu (or the area vector Va) to be updated by the adjacent area vector Va (or the user vector Vu) to update the user vector Vu or the area vector Va.

[0059] In this case, for example, the data of each dimension in the user vector Vu will be affected by the features of the adjacent area vector Va, and the data of each dimension in the area vector Va will be affected by the features of the adjacent user vector Vu.

[0060] On the other hand, in the example shown in Fig. 5(b), for example, after setting the user vector Vu and the area vector Va in separate dimensions so that they do not mix, the user vector Vu or the area vector Va is updated.

[0061] In this case, when updating a user vector having the user vector Vu1 and the area vector Va1 which is blank data using an area vector having the user vector Vu2 which is blank data and the area vector Va2, it is updated to a user vector having the user vector Vu1 and the area vector Va2.

[0062] That is, in this case, the features of the visited area are reflected in the area vector area of the user vector, and the features of the visited user are reflected in the user vector area of the area vector.

[0063] Further, the update unit 125 may update the area vector of each area using the area vectors of the areas adjacent to each area at a predetermined period (for example, once a day). Fig. 6 is a schematic diagram showing an example of updating the area vector according to the embodiment.

[0064] As shown in Fig. 6, since each area corresponds to the land in the real world, each node has a geographical relationship with each other. Therefore, for example, the area vectors are updated so that areas that are geographically close to each other have similar features.

[0065] For example, as shown in Fig. 6, in the area vectors Va11 to Va33 corresponding to the nine divided areas, the area vector Va22 located at the center is updated by the adjacent area vectors Va11 to Va21 and the area vectors Va23 to Va33.

[0066] That is, the update unit 125 updates the region vector so that the region vector Va22 and each region vector in the adjacent relationship are similar to each other. Thereby, the region vector can be appropriately updated.

[0067] Note that, for example, regarding the update of the region vector of each region, the update unit 125 may generate a graph in which adjacent nodes are connected by edges, and then update each region vector by GCNs for such a graph.

[0068] Returning to the description of FIG. 2, the provision unit 126 will be described. The provision unit 126 provides the user vector or the region vector updated by the update unit 125. For example, by providing the user vector to the portal site, the portal site can provide various services utilizing the user vector to the user.

[0069] Also, for example, the provision unit 126 can provide the region vector to facilities such as stores in the corresponding region, so that the facilities can carry out service deployment considering the characteristics of the region.

[0070] 〔3. Processing Procedure〕 Next, with reference to FIG. 7, the processing procedure executed by the update device 10 according to the embodiment will be described. FIG. 7 is a flowchart showing an example of the update process according to the embodiment.

[0071] As shown in FIG. 7, first, the update device 10 sets a bipartite graph with users and regions as nodes respectively (step S101). Subsequently, the update device connects the nodes (users and regions) based on the access history of the users (step S102).

[0072] Subsequently, the updating device 10 sets weights for the edges between the nodes (step S103). Then, the updating device 10 updates the vector (user vector or area vector) (step S104) and ends the process.

[0073] 〔4. Modification Example〕 Incidentally, in the above-described embodiment, the case where the user vector based on the actions of the user on the network and the area vector indicating the characteristics of the real-world area are combined based on the access history has been described. However, after combining the user and the products purchased by the user in the real world, the area can be extended to various objects, such as updating the user vector and the product vector indicating the characteristics of the product.

[0074] 〔5. Effects〕 The updating device 10 according to the above-described embodiment includes, in a bipartite graph including a set of users and a set of areas, a combining unit 123 that combines a user and the area visited by the user based on the access history of the user, and a user vector indicating the characteristics of the user based on the actions of the user on the network, and an updating unit 125 that updates the user vector or the area vector based on the area vector indicating the characteristics of the area combined with the user by the combining unit 123. Therefore, the updating device 10 according to the embodiment can update the user vector based on the area vector and can update the area vector based on the user vector, so that the vector indicating the characteristics can be appropriately updated.

[0075] In addition, the combining unit 123 according to the embodiment combines the areas according to the geographical relationship between the areas, and the updating unit 125 updates the area vectors of the respective areas using the weights according to the geographical relationship. Therefore, according to the updating device 10 according to the embodiment, the area vectors can be appropriately updated according to the geographical position relationship.

[0076] Further, the update unit 125 according to the embodiment updates the user vector or the area vector so that the user vector corresponding to the user adjacent to the area in the bipartite graph and the area vector are similar to each other. Therefore, according to the update device 10 according to the embodiment, each vector can be appropriately updated.

[0077] Further, the update device 10 according to the embodiment includes a setting unit 124 that sets a weight related to the update between the user and the area according to the relationship between the user and the area based on the access history. The update unit 125 updates the user vector or the area vector according to the weight related to the update set by the setting unit 124. Therefore, according to the update device 10 according to the embodiment, the weight can be appropriately set in the bipartite graph, so that each vector can be appropriately updated.

[0078] Further, the setting unit 124 according to the embodiment sets a weight related to the update based on the frequency of area access and the stay time by the user. Therefore, according to the update device 10 according to the embodiment, the weight related to the update can be appropriately set.

[0079] Further, the setting unit 124 according to the embodiment lowers the weight related to the update between the user and the area as the time of access by the user is older. Therefore, according to the update device 10 according to the embodiment, the bipartite graph can be appropriately updated, so that each vector can be appropriately updated.

[0080] Further, the update device 10 according to the embodiment includes a user vector generation unit 121 that generates a user vector of the user based on the purchase history, search history, and browsing history of the user on the network. Therefore, according to the update device 10 according to the embodiment, a user vector capturing the characteristics on the network of the user can be generated.

[0081] In addition, the user vector generation unit 121 according to the embodiment generates a user vector based on information regarding the user attributes of the user. Therefore, according to the update device 10 according to the embodiment, a more appropriate user vector can be generated.

[0082] In addition, the update device 10 according to the embodiment includes a region vector generation unit 122 that generates a region vector based on the user vector of the user who has visited the region. Therefore, according to the update device 10 according to the embodiment, a region vector considering the user vector of the user who has visited the region can be generated.

[0083] In addition, the region vector generation unit 122 according to the embodiment generates a region vector based on information regarding the facilities, roads, and sections of the region. Therefore, according to the update device 10 according to the embodiment, regions with more similar facilities and the like between regions can generate more similar region vectors.

[0084] In addition, the update method according to the embodiment is an update method executed by a computer. In a bipartite graph including a set of users and a set of regions, based on the access history of the users, a combination step of combining a user and the region visited by the user, and a user vector indicating the characteristics of the user based on the behavior of the user on the network, and a region vector indicating the characteristics of the region combined with the user by the combination step, and an update step of updating the user vector or the region vector. Therefore, according to the update method according to the embodiment, the vector indicating the characteristics can be appropriately updated.

[0085] In addition, the update program according to the embodiment causes a computer to execute an update procedure for updating a user vector or a region vector based on a combination procedure for combining a user and a region visited by the user in a bipartite graph including a set of users and a set of regions based on the user's access history, a user vector indicating characteristics of the user based on the user's actions on the network, and a region vector indicating characteristics of the region combined with the user by the combination procedure. Therefore, according to the update program according to the embodiment, the vector indicating the characteristics can be appropriately updated.

[0086] [6. Hardware Configuration] In addition, the update device 10 according to the above-described embodiment is realized by a computer 1000 having a configuration as shown in FIG. 8, for example. FIG. 8 is a hardware configuration diagram showing an example of a computer that realizes the functions of the update device 10 according to the embodiment. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM 1300, an HDD 1400, a communication interface (I / F) 1500, an input / output interface (I / F) 1600, and a media interface (I / F) 1700.

[0087] The CPU 1100 operates based on a program stored in the ROM 1300 or the HDD 1400 and controls each part. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, a program dependent on the hardware of the computer 1000, and the like.

[0088] The HDD 1400 stores a program executed by the CPU 1100, data used by such a program, and the like. The communication interface 1500 receives data from other devices via a network (communication network) N and sends it to the CPU 1100, and transmits data generated by the CPU 1100 to other devices via the network N.

[0089] The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice (in FIG. 8, output devices and input devices are collectively referred to as "input / output devices") via the input / output interface 1600. The CPU 1100 acquires data from the input device via the input / output interface 1600. Further, the CPU 1100 outputs the generated data to the output device via the input / output interface 1600.

[0090] The media interface 1700 reads a program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1200. The CPU 1100 loads such a program from the recording medium 1800 onto the RAM 1200 via the media interface 1700 and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase change rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory.

[0091] For example, when the computer 1000 functions as the update device 10 according to the embodiment, the CPU 1100 of the computer 1000 realizes the functions of the control unit 120 by executing the program loaded on the RAM 1200. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800. As another example, these programs may be acquired from another device via the network N.

[0092] As described above in detail some of the embodiments of the present application with reference to the drawings, these are examples, and the present invention can be implemented in other forms with various modifications and improvements based on the knowledge of those skilled in the art, starting from the aspects described in the column of the disclosure of the invention.

[0093] 〔7. Others〕 Also, among the respective processes described in the above embodiments and modification examples, all or part of the processes described as being automatically performed can be manually performed, or all or part of the processes described as being manually performed can be automatically performed by a known method. In addition, regarding the processing procedures, specific names, and information including various data and parameters shown in the above documents and drawings, they can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the illustrated information.

[0094] Also, each component of each device shown in the drawings is a functional concept, and it is not necessarily physically configured as shown in the drawings. That is, the specific form of the distribution and integration of each device is not limited to that shown in the drawings, and all or part of it can be functionally or physically distributed and integrated in arbitrary units according to various loads and usage situations.

[0095] Also, the above-described embodiments and modification examples can be appropriately combined within a range that does not conflict with the processing content.

[0096] Also, the above-described "section (section, module, unit)" can be read as "means", "circuit", etc. For example, the acquisition unit can be read as an acquisition means or an acquisition circuit.

Description of Reference Numerals

[0097] 10 Updating device 110 Communication unit 120 Control unit 121 User vector generation unit 122 Region vector generation unit 123 Combining unit 124 Setting unit 125 Updating unit 126 Providing unit 130 Storage unit 131 User vector model storage unit 132 User vector storage unit 133 Region vector model storage unit 134 Region vector memory unit 135 Map information memory unit 136 Bipartite graph memory unit Va Region vector Vu User vector

Claims

1. In a bipartite graph including a set of users and a set of regions, a combining unit that combines a user and a region visited by the user based on the user's access history in the real world, and an updating unit that updates the user vector or the region vector based on a user vector indicating characteristics of the user based on the user's behavior on the network and a region vector indicating characteristics of the region combined with the user by the combining unit. An updating device characterized by comprising the above.

2. The combining unit combines the regions according to the geographical relationship between the regions, and the updating unit updates the region vectors of the respective regions using weights according to the geographical relationship. The updating device according to claim 1, characterized by the above.

3. The updating unit updates the user vector or the region vector so that the user vector and the region vector corresponding to the user and the region adjacent to each other in the bipartite graph are similar to each other. The updating device according to claim 1 or 2, characterized by the above.

4. Comprising a setting unit that sets a weight related to the update between the user and the region according to the relationship between the user and the region based on the access history, and the updating unit updates the user vector or the region vector according to the weight related to the update set by the setting unit. The updating device according to claim 1, 2 or 3, characterized by the above.

5. The setting unit sets the weight related to the update based on the access frequency and stay time of the region by the user. The updating device according to claim 4, characterized by the above.

6. The setting unit lowers the weight related to the update between the user and the region as the time of access by the user is older. The updating device according to claim 4 or 5, characterized by the above.

7. A user vector generation unit that generates the user vector of the user based on the purchase history, search history, and browsing history of the user on the network. The updating device according to any one of claims 1 to 6, characterized by comprising the above.

8. The user vector generation unit generates the user vector based on information related to the user attributes of the user. The updating device according to claim 7, characterized by the above.

9. An area vector generation unit that generates the area vector based on the user vector of the user who visited the area The update device according to any one of claims 1 to 8, characterized by comprising the same

10. The area vector generation unit Generates the area vector based on information regarding facilities, roads, and sections of the area The update device according to claim 9, characterized by the same

11. An update method executed by a computer, comprising In a bipartite graph including a set of users and a set of areas, a combining step of combining the user and the area visited by the user based on the user's visit history in the real world An update step of updating the user vector or the area vector based on a user vector indicating characteristics of the user based on the user's behavior on the network and an area vector indicating characteristics of the area coupled to the user by the combining step The update method characterized by including the same

12. In a bipartite graph including a set of users and a set of areas, a combining procedure of combining the user and the area visited by the user based on the user's visit history in the real world An update procedure of updating the user vector or the area vector based on a user vector indicating characteristics of the user based on the user's behavior on the network and an area vector indicating characteristics of the area coupled to the user by the combining procedure An update program characterized by causing a computer to execute the same

Images