People flow prediction method, apparatus, and program

Abstract

To achieve appropriate people flow prediction considering short-term variable elements.SOLUTION: A people flow prediction method predicts a people flow in a prediction target area in a prediction target period on the basis of input data. The input data includes people flow big data (D1) including attribute data of each user and movement history data of each user, event history data (D6) representing the presence or absence of execution of an event and the actual result indicating the details of the executed event in the prediction target area before the prediction target period, and event schedule data representing the presence or absence of execution of an event and the schedule of the details of the execution in the prediction target period and prediction target area.SELECTED DRAWING: Figure 6

Description

[Technical field]

[0001] The present invention relates to a people flow prediction method, device, and program. [Background technology]

[0002] When the introduction of a new large-scale facility or transportation system is considered over a period of several years, a simulator related to people flow or traffic flow is used as a means of predicting the effects of the introduction (EBPM: Evidence Based Policy Making). The simulator described here is called the reference simulator. Reference simulators are mainly used by government agencies or transportation operators.

[0003] When predicting the effect of introducing a new large-scale facility or transportation system in the reference simulator, the person trip data of visitors to the city area and the tendency of visitors to choose facilities and transportation methods that they usually use are referenced. The person trip data of visitors corresponds to the attribute data of visitors. The tendency of choice is identified from information collected by questionnaires, etc. In the reference simulator, the person trip data and the tendency of choice are used to model the behavior of visitors, and the effect of introduction is predicted by calculating the probability that visitors will choose the new large-scale facility or transportation system. In addition, a technology has been proposed for predicting people flow by referring to weather conditions, greenery view rate, scenery, buildings, stores, store attractiveness, transportation convenience, etc. (see

[0020] of Patent Document 1 below). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2022-88278 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the above person trip data and choice trends are data collected by government agencies over a short period of time, once every few years, and therefore the reference simulator cannot make predictions that take into account short-term fluctuations.

[0006] The present invention aims to provide a people flow prediction method, device, and program that can perform appropriate people flow prediction taking into account short-term variable factors. [Means for solving the problem]

[0007] The people flow prediction method of the present invention is a people flow prediction method that predicts people flow within a prediction target area during a prediction period based on input data, wherein the input data includes people flow big data including attribute data of each user and movement history data of each user, event history data indicating whether or not an event has been held in the prediction target area prior to the prediction period and the actual content of the event, and event schedule data indicating whether or not an event will be held during the prediction period and in the prediction target area, and the schedule of the event. Effect of the Invention

[0008] According to the present invention, it is possible to provide a people flow prediction method, device, and program capable of making an appropriate people flow prediction that takes into account short-term variable factors. [Brief description of the drawings]

[0009] [Figure 1] 1 is an overall configuration diagram of a system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram illustrating an internal configuration of a terminal device according to an embodiment of the present invention. [Diagram 3] 2 is a diagram illustrating an internal configuration of a service providing device according to an embodiment of the present invention. [Figure 4] 2 is a diagram showing an internal configuration of a service use device according to an embodiment of the present invention. [Diagram 5] FIG. 2 is a diagram showing a correspondence relationship between a plurality of terminal devices and a plurality of users in the embodiment of the present invention. [Figure 6]FIG. 2 is an explanatory diagram of data stored in a database according to an embodiment of the present invention. [Figure 7] FIG. 2 is a diagram showing a structure of people flow big data according to an embodiment of the present invention. [Figure 8] FIG. 2 is a diagram illustrating the generation and acquisition timing of people flow big data according to an embodiment of the present invention. [Figure 9] 4 is an operation flowchart of the terminal device 10 according to the embodiment of the present invention, which follows a travel assistance application program. [Figure 10] FIG. 4 is a diagram showing a structure of application registration data according to the embodiment of the present invention. [Figure 11] FIG. 4 is a diagram showing a structure of application log data in the embodiment of the present invention. [Figure 12] FIG. 4 is an explanatory diagram of input / output data of a simulator according to the embodiment of the present invention. [Figure 13] 5 is an operation flowchart of the service providing device according to the embodiment of the present invention. [Figure 14] FIG. 11 is a diagram showing the relationship between a prediction time point of a simulator, a prediction target period, and a reference period according to an embodiment of the present invention. [Figure 15] FIG. 11 is a diagram showing the relationship between four assumed periods in a first example belonging to an embodiment of the present invention. [Figure 16] FIG. 11 is a diagram showing a state in which shopping facilities exist in an area of ​​interest according to a second embodiment of the present invention. [Figure 17] FIG. 11 is a diagram showing the relationship between four assumed periods in a second example belonging to an embodiment of the present invention. [Figure 18] FIG. 11 is a diagram showing actual visitor data (a)-(c) of a shopping facility on the first to third dates and predicted visitor data (d) of the shopping facility on a predicted date in a second example pertaining to an embodiment of the present invention. [Figure 19] FIG. 13 is a diagram showing a flow of processing that can be executed by a simulator according to a ninth example pertaining to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, examples of the embodiments of the present invention will be described in detail with reference to the drawings. In each of the drawings, the same parts are given the same reference numerals, and duplicated descriptions of the same parts are omitted as a rule. In this specification, for the sake of simplifying the description, a symbol or code referring to information, a signal, a physical quantity, a functional part, a circuit, an element, or a part, etc. may be written, and the name of the information, signal, physical quantity, functional part, circuit, element, or part, etc. corresponding to the symbol or code may be omitted or abbreviated. For example, the people flow big data referred to by "D1" described later (see FIG. 6) may be written as people flow big data D1 or abbreviated as data D1, but they all refer to the same thing.

[0011] FIG. 1 shows the overall configuration of a system SYS according to an embodiment of the present invention. The system SYS is a service providing system having at least a service providing device 20 and a database 30 as components. The service providing device 20 provides one or more services including a people flow prediction service while referring to the data in the database 30. In the people flow prediction service, the service providing device 20 simulates people flow within a prediction target area and outputs the simulation results. In the people flow prediction service, the movement behavior of each person within the prediction target area is predicted. Therefore, the simulation of people flow is, in other words, a simulation of the movement behavior of each person.

[0012] FIG. 1 shows a plurality of terminal devices 10. Each terminal device 10 is an electronic device that is held and operated by an operator of the terminal device 10. For example, a smartphone, a tablet, a personal computer, or the like can be used as the terminal device 10. Each terminal device 10 may be understood as not being included as a component of the system SYS, or as being included as a component of the system SYS. The operator of the terminal device 10 functions as a user of the system SYS, and hereinafter, a user refers to a person who holds and operates any of the terminal devices 10.

[0013] FIG. 1 also shows a service usage device 40. The service usage device 40 is a device that uses a service provided by the service providing device 20. Strictly speaking, it is the operator of the service usage device 40 who receives the provision of the service. In FIG. 1, the service usage device 40 is shown separately from a group of multiple terminal devices 10, but any of the terminal devices 10 can be the service usage device 40. In this case, the operator of any of the terminal devices 10 becomes the operator of the service usage device 40. The results of the people flow simulation performed by the service providing device 20 are output from the service providing device 20 to the service usage device 40, and the simulation results are transmitted to the operator.

[0014] 1 also shows a provider-side server device 50. The provider-side server device 50 is a server device operated by a telecommunications carrier. The service usage device 40 and the provider-side server device 50 are not included in the components of the system SYS. However, it is also possible to interpret that at least one of the service usage device 40 and the provider-side server device 50 is included in the components of the system SYS.

[0015] Each of the terminal devices 10, the service providing device 20, the service using device 40, and the provider side server device 50 is connected to a predetermined communication network NET by wire or wirelessly. Since each of the terminal devices 10 is assumed to be carried by the corresponding user while moving, each of the terminal devices 10 is in principle connected to the communication network NET wirelessly. Any device connected to the communication network NET (i.e., devices 10, 20, 40, and 50) can communicate bidirectionally with any other device connected to the communication network NET via the communication network NET. Although not clear from FIG. 1, the database 30 may also be connected to the communication network NET.

[0016] The communication network NET includes all or part of a mobile communication network including base stations, the Internet, a wireless LAN (Local Area Network), and a short-distance wireless communication line. The wireless LAN may be, for example, compliant with Wi-Fi (registered trademark). The short-distance wireless communication line may be, for example, compliant with Bluetooth (registered trademark).

[0017] FIG. 2 shows a schematic block diagram of one terminal device 10. In the system SYS, any two terminal devices 10 may be different types of terminal devices. That is, for example, one terminal device 10 may be a smartphone and the other terminal device 10 may be a personal computer. Alternatively, any two terminal devices 10 may be different types of smartphones. However, in the following description, each terminal device 10 is assumed to have the configuration shown in FIG. 2. That is, each terminal device 10 in the system SYS is assumed to include a controller 11, a memory 12, a communication unit 13, a GPS processing unit 14, and an interface unit 15.

[0018] The controller 11 includes, as hardware resources, a calculation processing unit including a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), etc. The controller 11 may realize each of the following functions to be realized by the controller 11 by executing a program recorded in the memory 12 or any other recording medium (not shown).

[0019] The memory 12 is configured to include a non-volatile memory such as a ROM (Read only memory) or a flash memory, and a volatile memory such as a RAM (Random access memory). The memory 12 stores various data referenced by the controller 11, as well as various programs to be executed by the controller 11.

[0020] The communication unit 13 transmits and receives any information between the terminal device 10 and a counterpart device different from the terminal device 10. The counterpart device for the communication unit 13 includes any device connected to the communication network NET. Note that the controller 11 can transmit and receive any information to and from the counterpart device (the counterpart device for the communication unit 13) using the communication unit 13, but in the following, the description of the communication unit 13 may be omitted.

[0021] The GPS processing unit 14 detects the position of the terminal device 10 (the current location of the terminal device 10) by receiving signals from multiple GPS satellites that form a GPS (Global Positioning System). The detected position of the terminal device 10 is referred to as the terminal position, and information indicating the terminal position is referred to as terminal position information. The terminal position information is sent from the GPS processing unit 14 to the controller 11. The terminal position of a certain terminal device 10 may be considered to represent the current location (location at the current time) of the user of the terminal device 10.

[0022] The interface unit 15 is a man-machine interface between the terminal device 10 and the operator of the terminal device 10 (i.e., the user of the terminal device 10), and includes a display screen 15a, an operation unit 15b, a microphone 15c, and a speaker 15d. The display screen 15a is configured with a liquid crystal display panel or the like, and displays any image under the control of the controller 11. The operation unit 15b accepts any operation from the operator of the terminal device 10. The display screen 15a and the operation unit 15b may form a touch panel, and the operation on the operation unit 15b may be an operation on the touch panel. The microphone 15c collects sounds around the microphone 15c and converts the collected sounds into electrical signals. The speaker 15d outputs any sound under the control of the controller 11.

[0023] 3 shows a schematic block diagram of the service providing device 20. The service providing device 20 includes a controller 21, a memory 22, and a communication unit 23. The service using device 40 may be configured with one or more computer devices connected to a communication network NET. The service using device 40 may be configured using cloud computing.

[0024] The controller 21 includes a calculation processing unit including a CPU, a GPU, etc. as hardware resources, and has function blocks F1 to F4 as functional blocks configured by the calculation processing unit. The controller 21 may realize the following functions (including the functions of the function blocks F1 to F4) to be realized by the controller 21 by executing a program recorded in the memory 22 or any other recording medium (not shown). The function blocks F1, F2, F3, and F4 are a data acquisition unit, a simulator, a result output unit, and an application linking unit, respectively. Details of these function blocks will be described later.

[0025] The memory 22 is configured to have a non-volatile memory such as a ROM or a flash memory, and a volatile memory such as a RAM. The memory 22 stores various data referenced by the controller 21, as well as various programs to be executed by the controller 21. The communication unit 23 transmits and receives any information between the service providing device 20 and a counterpart device different from the service providing device 20. The counterpart device for the communication unit 23 includes any device connected to the communication network NET. The controller 21 can transmit and receive any information to and from the counterpart device (the counterpart device for the communication unit 23) using the communication unit 23, but the description of the communication unit 23 may be omitted below.

[0026] 4 shows a schematic block diagram of service usage device 40. Service usage device 40 has the same configuration as terminal device 10, and includes a controller 41, a memory 42, a communication unit 43, and an interface unit 45. Service usage device 40 may include a GPS processing unit equivalent to GPS processing unit 14 in FIG.

[0027] The controller 41 includes a calculation processing unit including a CPU and a GPU as hardware resources. The controller 41 may execute a program recorded in the memory 42 or any other recording medium (not shown) to realize the following functions to be realized by the controller 41. The memory 42 is configured to have a non-volatile memory such as a ROM or a flash memory and a volatile memory such as a RAM. The memory 42 stores various data referenced by the controller 41 as well as various programs to be executed by the controller 41. The communication unit 43 transmits and receives any information between the service using device 40 and a counterpart device different from the service using device 40. The counterpart device for the communication unit 43 includes any device connected to the communication network NET. The controller 41 can transmit and receive any information to and from the counterpart device (the counterpart device for the communication unit 43) using the communication unit 43, but the description of the communication unit 43 may be omitted below.

[0028] The interface unit 45 is a man-machine interface between the service utilization device 40 and the operator of the service utilization device 40, and includes a display screen 45a, an operation unit 45b, a microphone 45c, and a speaker 45d. The display screen 45a is configured with a liquid crystal display panel or the like, and displays any image under the control of the controller 41. The operation unit 45b accepts any operation from the operator of the service utilization device 40. The display screen 45a and the operation unit 45b may form a touch panel, and the operation on the operation unit 45b may be an operation on the touch panel. The microphone 45c collects sounds around the microphone 45c and converts the collected sounds into electrical signals. The speaker 45d outputs any sound under the control of the controller 41.

[0029] Although not particularly shown, the provider-side server device 50 has a hardware configuration equivalent to that of the service providing device 20 shown in Fig. 3. The provider-side server device 50 may be configured with one or more computer devices connected to a communication network NET. The provider-side server device 50 may also be configured using cloud computing.

[0030] With reference to FIG. 5, in the following, when it is necessary to distinguish between multiple terminal devices 10, the multiple terminal devices 10 are referred to as terminal devices 10[1] to 10[N]. N represents an integer of 2 or more, and is basically sufficiently larger than 2 (for example, tens of thousands or more). A user of terminal device 10[i] is particularly referred to as user U[i]. i represents any integer.

[0031] [Database] 6 shows a group of data stored in the database 30. The database 30 may be composed of a plurality of databases installed in different locations. Any data stored in the database 30 is acquired by a data acquisition unit F1, and the data acquired by the data acquisition unit F1 is stored in the database 30. However, some of the data stored in the database 30 may be acquired by an application linkage unit F4.

[0032] People flow big data D1 is stored in the database 30. The people flow big data D1 is collected and generated by the business-side server device 50, and the data acquisition unit F1 acquires the people flow big data D1 from the business-side server device 50.

[0033] FIG. 7 shows the structure of people flow big data D1. The people flow big data D1 has user unit data for each user. The user unit data has attribute data and movement history data. Any one of the noticed users is called a noticed user, and the user unit data of the noticed user will be described.

[0034] The attribute data of the noted user represents attributes of the noted user, and includes data representing the noted user's age group, sex, nationality, address, occupation, and preferences.

[0035] When the user of interest is user U[i], user U[i] inputs his / her own attribute data to terminal device 10[i]. The input attribute data is transmitted from terminal device 10[i] to provider-side server device 50, whereby provider-side server device 50 collects the attribute data of user U[i].

[0036] The movement history data of the attention user represents the location and movement history of the attention user. The movement history data of the attention user specifies when the attention user visited which location, by which means of transportation, and how long the attention user stayed at the location. Therefore, for example, when considering a visit to a certain attention location, the movement history data of the attention user represents the date and time when the attention user visited the attention location and the means of transportation used when the attention user visited the attention location. Furthermore, the movement history data of the attention user also represents unique information of the attention location (location, type, etc.) and the attention user's stay time at the attention location. The attention location may be a store selling products, a movie theater, a concert venue, a sports facility, a public facility, etc.

[0037] When the user of interest is user U[i], the terminal location information acquired by the terminal device 10[i] is transmitted to the provider-side server device 50, whereby the location of the terminal device 10[i] is recognized by the provider-side server device 50. The location of the terminal device 10[i] is regarded as the location of the user U[i]. The provider-side server device 50 may acquire the location of the terminal device 10[i] by base station positioning instead of GPS positioning. The provider-side server device 50 estimates the means of transportation used by the user U[i], the visited points of the user U[i], and the time spent at the visited points from the time-series data of the location of the terminal device 10[i], with reference to map data as well. Based on the estimation result, the provider-side server device 50 generates the movement history data of the user U[i]. The attribute data and movement history data of the user of interest are stored in the database 30 in a manner in which the user of interest cannot be identified (individual cannot be identified).

[0038] The data acquisition unit F1 can acquire people flow big data D1 including movement history data from a certain period of time ago from the current time. A further explanation will be given on this. With reference to FIG. 8, the current time is assumed to be time 610. Time 612 is a certain period of time 611 ago from the current time 610. The business-side server device 50 sequentially generates new movement history data as time passes. However, the people flow big data D1 that can be provided from the business-side server device 50 to the service providing device 20 at time 610 is the people flow big data D1 including movement history data before time 612. The people flow big data D1 including movement history data after time 612 at time 610 is not provided from the business-side server device 50 to the service providing device 20.

[0039] The points in time 610 and 612 may be points in time in days. In this case, for example, the fixed period 611 is seven days (i.e., one week), and the business-side server device 50 may generate new movement history data in days. In other words, the business-side server device 50 generates new movement history data in days, and includes only movement history data from one week or more before the current point in time 610 in the movement history data of the people flow big data D1.

[0040] For example, if the current time 610 is September 20, 2023, the people flow big data D1 including movement history data before September 13, 2023 can be provided from the business-side server device 50 to the service providing device 20. On September 20, 2023, the data acquisition unit F1 can acquire the people flow big data D1 including movement history data before September 13, 2023 and store it in the database 30. On September 20, 2023, the data acquisition unit F1 cannot acquire the people flow big data D1 including movement history data after September 13, 2023. That is, for example, at 9:00 a.m. on September 20, 2023, the movement history data from 0:00 a.m. on September 14, 2023 to just before 9:00 a.m. on September 20, 2023 is not acquired by the data acquisition unit F1 and is not included in the people flow big data D1. Note that one week given as the fixed period 611 is merely an example, and the fixed period 611 may be other than one week. The points in time 610 and 612 may be understood to be two different times, and the fixed period 611 may be a period of any length of time (for example, 11 hours or 48 hours).

[0041] Map data D2 is stored in database 30. Map data D2 represents a map of the service provision area. In this embodiment, the service provision area is assumed to be the entire country of Japan. Information indicating the locations and types of facilities, roads, and transportation networks in the service provision area is included in map data D2. Data acquisition unit F1 can acquire existing map data as map data D2. There are many facilities in the service provision area. The facilities in the service provision area include stores selling products, movie theaters, concert venues, sports facilities, public facilities, etc.

[0042] Operation-related data D3 is stored in the database 30. The operation-related data D3 includes operation schedule data D3a and congestion status data D3b. The operation schedule data D3a represents the operation schedule of transportation facilities located within the service provision area. The transportation facilities here refer to any transportation facilities operating within the service provision area, such as trains, buses, airplanes, aircraft, and ships. Trains include electric trains (including subway trains) and steam locomotives. The operation schedule data D3a indicates the operation plan of each transportation facility. The congestion status data D3b represents the congestion status of each transportation facility.

[0043] The bus schedule data D3a includes at least basic schedule data D3a_1. The basic schedule data D3a_1 indicates a basic operation plan for each transportation facility, and in principle, each transportation facility operates according to the basic schedule data D3a_1. A transportation facility operator may set a special operation period in accordance with some demand, and in this case, the corresponding transportation facility operates temporarily. The operation plan of the transportation facility during the special operation period is indicated by special schedule data D3a_2, and the transportation facility operating temporarily operates according to the special schedule data D3a_2. When a special operation period is set, the special schedule data D3a_2 is included in the bus schedule data D3a. The special schedule data D3a_2 may be set after the basic schedule data D3a_1 is set. In other words, a transportation facility operator may first set the basic schedule data D3a_1, and then set the special schedule data D3a_2 temporarily. The data acquisition unit F1 is a server device connected to the communication network NET and can acquire bus schedule data D3a and congestion status data D3b from a server device (not shown) operated by a transportation operator.

[0044] In addition, a temporary schedule may be set suddenly due to a sudden traffic disturbance caused by bad weather or an accident. In this case, the data acquisition unit F1 can acquire temporary schedule data D3a_2 according to the latest operation status. The temporary schedule data D3a_2 of a certain transportation facility may only indicate that the transportation facility is stopped.

[0045] Weather history data D4 and weather forecast data D5 are stored in database 30. Weather history data D4 represents the history of weather at each point within the service area. Weather forecast data D5 represents the weather forecast (forecasted weather) at each point within the service area. Data acquisition unit F1 can acquire weather history data D4 and weather forecast data D5 from a weather information providing server device (not shown) connected to communication network NET.

[0046] Event history data D6 is stored in the database 30. The event history data D6 represents the history of events performed at any point in the service area. Events include events that are held periodically or irregularly. For example, concerts held at concert venues, athletic matches held at athletic facilities, and various festivals are events. For example, sales held at stores that sell products are also events. Sales are sales that involve discounts on products, and coupons may be issued to take advantage of the discounts. Sales that involve the issuance of coupons are also a type of event. For any target store, a coupon for the target store is a benefit that can be used at the target store. A coupon for the target store is a benefit such as a gift certificate or discount coupon that can be used at the target store, or a gift certificate for an item such as a commemorative item, or a participation ticket for an event, etc. A coupon may be issued to the terminal device 10, and in this case, the coupon is a virtual (electronic) gift certificate or discount coupon that is referenced on the terminal device 10.

[0047] When an event is held at any point within the service area, the data acquisition unit F1 acquires event record data indicating the details of the event, and includes the data in the event history data D6. The event record data specifies the details of the event, such as the date and time of the event, the location of the event, the number of participants, and the type of the event. The event record data is acquired for each event and included in the event history data D6. A server device (not shown) for managing events is connected to the communication network NET, and the event record data is generated and collected by the server device for managing events.

[0048] For example, when a concert is held as an event, information such as the date and time, location, number of participants, and type of the concert is input to the event management server device by an operator or the like directly or via a computer device. As a result, event record data for the concert is generated in the event management server device. The event management server device may automatically extract information about the concert disclosed on an information and communication network such as the Internet, thereby generating event record data for the concert. Similarly, for example, when a sale is held at a certain target store, information such as the date and time, location, and type of the sale is input to the event management server device by an operator or the like directly or via a computer device. As a result, event record data for the sale is generated in the event management server device. The event management server device may automatically extract information about the sale disclosed on an information and communication network such as the Internet, thereby generating event record data for the sale.

[0049] The data acquisition unit F1 can acquire event history data from a server device for event management, but the data acquisition unit F1 itself may generate and collect the event history data.

[0050] The application registration data D11 and the application log data D12 are stored in the database 30. The application registration data D11 and the application log data D12 are acquired using a movement assistance application program. The movement assistance application program is hereinafter referred to as a movement assistance application. The movement assistance application is installed in all or some of the terminal devices 10 among the multiple terminal devices 10 operated in the system SYS. In the terminal device 10 in which the movement assistance application is installed, the controller 11 executes the movement assistance application. Unless otherwise specified, the operation of the terminal device 10 described below may be understood as an operation performed in accordance with the movement assistance application by the controller 11 executing the movement assistance application. Furthermore, hereinafter, unless otherwise specified, it is assumed that the movement assistance application is installed in any terminal device 10.

[0051] With reference to FIG. 9, the operation of the terminal device 10[i] according to the migration support application will be described. FIG. 9 is an operation flowchart of the terminal device 10[i] according to the migration support application. When the controller 11 in the terminal device 10[i] executes the migration support application for the first time, the controller 11 executes the user registration process of step S11. In the user registration process in the terminal device 10[i], the user U[i] inputs user registration data to the terminal device 10[i] according to the inquiry items displayed on the display screen 15a. The user registration data of the user U[i] includes attribute data of the user U[i] (see FIG. 10), and the attribute data of the user U[i] includes data representing the age group, sex, nationality, address, occupation, preferences, etc. of the user U[i]. The user registration data of the user U[i] further includes account information of the user U[i] (see FIG. 10).

[0052] In step S12 following step 11, the controller 11 transmits the input user registration data to the service providing device 20. Thereafter, the controller 11 repeatedly executes a terminal location information transmission process in step S13 and a distribution response process in step S14. In the terminal location information transmission process, the latest terminal location information generated by the GPS processing unit 14 is transmitted from the terminal device 10[i] to the service providing device 20. The distribution response process will be described later.

[0053] The application cooperation unit F4 (see FIG. 3) operates in cooperation with the movement assistance app. The application cooperation unit F4 stores the user registration data sent from each terminal device 10 in application registration data D11 (see FIG. 6) and stores it in the database 30. FIG. 10 shows the structure of the application registration data D11. In the application registration data D11, unique identification information (ID) is assigned to each user who uses the movement assistance app. In the application registration data D11, user registration data is stored in association with the identification information (ID) for each user who uses the movement assistance app. When the movement assistance app is installed in the terminal device 10[i], a user U[i] corresponds to a user who uses the movement assistance app.

[0054] 11 shows the structure of the application log data D12. In the application log data D12, location data, action log data, and movement support log data are stored in association with identification information (ID) for each user who uses the movement support app. Assuming that the movement support app is installed in the terminal device 10[i] and that the user who uses the movement support app is user U[i], the application log data D12 for user U[i] will be described.

[0055] The controller 11 of the terminal device 10[i] transmits the terminal location information to the service providing device 20. In the terminal device 10[i], the GPS processing unit 14 generates and updates the terminal location information at a predetermined cycle (for example, at intervals of several seconds). The controller 11 in step S13 transmits the latest terminal location information generated by the GPS processing unit 14 to the service providing device 20 in real time. Here, real-time transmission refers to transmitting the latest terminal location information with a sufficiently short delay compared with the length of the above-mentioned fixed period 611 (see FIG. 8). That is, the terminal location information generated at the first time is transmitted to the service providing device 20 at the second time, and the time difference between the first time and the second time is sufficiently shorter than the fixed period 611, for example, several seconds to several tens of seconds.

[0056] The application cooperation unit F4 stores location data based on the terminal location information received from the terminal device 10[i] in association with the user U[i] in the application log data D12. The location data corresponding to the user U[i] is the latest terminal location information received from the terminal device 10[i], and represents the current location of the user U[i] (however, there may be an error). That is, the terminal location information of the terminal device 10[i] is sequentially stored in real time in the application log data D12 as the location data of the user U[i]. Here, storing in real time means storing the latest location data with a sufficiently short delay compared with the length of the above-mentioned fixed period 611 (see FIG. 8). That is, the terminal location information of the terminal device 10[i] generated at the first time is stored in the application log data D12 as the location data of the user U[i] substantially at the second time. The time difference between the first time and the second time is sufficiently shorter than the fixed period 611 (see FIG. 8), for example, several seconds to several tens of seconds.

[0057] The application linking unit F4 generates action log data of the user U[i] based on the terminal location information sequentially received from the terminal device 10[i]. The application linking unit F4 associates the action log data of the user U[i] with the user U[i] and stores it in the application log data D12. The action log data of the user U[i] indicates a time-series change in the location of the user U[i]. The application linking unit F4 can estimate the means of transportation used by the user U[i], the places visited by the user U[i], and the time spent at the places visited by the user U[i] based on the terminal location information sequentially received from the terminal device 10[i], with reference to the map data D2. The application linking unit F4 may include the estimation result in the action log data of the user U[i].

[0058] The application linking unit F4 provides navigation support to the user U[i] based on the map data D2 and the user registration data, location data, and action log data of the user U[i]. The navigation support log data for the user U[i] is log data related to navigation support for the user U[i].

[0059] In the wandering support, the application cooperation unit F4 distributes (in other words, transmits) wandering guidance information to the terminal device 10[i]. The wandering guidance information is, for example, information that proposes a visit to a stopover suggested facility located near the current location of the user U[i]. When the terminal device 10[i] receives the visit-off suggested information, the controller 11 notifies the user U[i] by displaying the visit-off suggested information on the display screen 45a. The notification is included in the delivery response process of step S14 (see FIG. 9). The stop-off suggested facility may be any facility including a store, or may be a tourist spot, etc. The application cooperation unit F4 can set the stop-off suggested facility based on information about the preferences of the user U[i]. In addition, the wandering guidance information may be any information that supports the wandering of the user U[i]. When the application cooperation unit F4 distributes the wandering guidance information to the terminal device 10[i], it includes data indicating that (delivery history data of the wandering guidance information) together with the delivery date and time in the wandering support log data related to the user U[i].

[0060] In the wandering support, when the application cooperation unit F4 distributes the stop-off guide information to the terminal device 10[i], it may distribute to the terminal device 10[i] a coupon that can be used at the suggested stop-off facility. When the application cooperation unit F4 distributes the coupon to the terminal device 10[i], it includes data indicating that (coupon distribution history data) together with the distribution date and time in the wandering support log data for the user U[i].

[0061] After the terminal device 10[i] receives a coupon that can be used at the suggested stop-off facility, the user U[i] can actually use the coupon at the suggested stop-off facility. When the user U[i] actually uses the coupon, the application linking unit F4 includes data indicating that (coupon use history data) together with the date and time of use in the travel support log data for the user U[i]. For example, the user U[i] can use the coupon by having an information terminal (not shown) installed at the suggested stop-off facility read the code information of the coupon displayed on the terminal device 10[i]. At this time, the fact that the coupon has been used is transmitted to the service providing device 20 directly from the information terminal or via any other server device (not shown).

[0062] [Simulator] Here, we will explain matters related to simulators. When the introduction of a new large-scale facility or transportation system is considered over a period of several years, simulators related to people flow or traffic flow are used as a means of predicting the effects of the introduction (EBPM: Evidence Based Policy Making). The simulator described here is a reference simulator different from simulator F2 in Figure 3. Reference simulators are mainly used by government agencies or transportation operators.

[0063] When predicting the effect of introducing a new large-scale facility or transportation system in the reference simulator, the person trip data of visitors visiting the city area and the selection tendency of the facility and transportation means that the visitors usually use are referenced. The person trip data of the visitors corresponds to the attribute data of the visitors. The selection tendency is identified from information collected by questionnaires, etc. In the reference simulator, the person trip data and the selection tendency are used to model the behavior of the visitors, and the introduction effect is predicted by calculating the probability that the visitors will select the new large-scale facility or transportation system. However, the above person trip data and the selection tendency are data obtained by administrative agencies for a short period of time once every few years. Therefore, the reference simulator cannot make predictions that take into account the influence of the season to which the prediction period belongs, the weather in the predicted area, special transportation schedules, and events to be held. Therefore, it is difficult to apply the reference simulator to verify the effect of visitor attraction measures planned daily by each facility operator.

[0064] Meanwhile, in recent years, telecommunications carriers have begun to provide people flow big data that includes movement history data of each person in addition to data equivalent to the person trip data described above, making it possible to collect data throughout the year. The simulator F2 according to this embodiment precisely predicts people flow and the movement behavior of visitors in any area by referring to various other data together with the above-mentioned people flow big data D1 (see Figs. 6 and 7).

[0065] The simulator F2 predicts the flow of people (in other words, the movement behavior of visitors to the area) in any area within the service provision area. The area to be predicted is referred to as the prediction target area. The period to be predicted is referred to as the prediction target period. Furthermore, data referenced for prediction is referred to as the prediction input data IN.

[0066] Fig. 12(a) shows input / output data of simulator F12. An operator of the service usage device 40 inputs prediction condition data to the service usage device 40. The prediction condition data is transmitted from the service usage device 40 to the service providing device 20, and the prediction condition data is included in the input data IN for prediction. As shown in Fig. 12(b), the prediction condition data includes data specifying at least the prediction target area and the prediction target period. The prediction condition data may further include additional condition data. Details will be described later, but the additional condition data is, for example, event schedule data indicating the schedule for holding an event.

[0067] The simulator F2 performs people flow prediction processing based on the input data IN for prediction. In the people flow prediction processing based on the input data IN for prediction, the simulator F2 predicts people flow (in other words, the movement behavior of visitors to the area) within the prediction target area during the prediction target period, and generates prediction result data OUT indicating the result of the prediction. The result output unit F3 transmits the prediction result data OUT to the service usage device 40. In the service usage device 40, the controller 41 displays the received prediction result data OUT on the display screen 45a.

[0068] The input data IN for prediction includes all data (hence, data D1 to D6, D11 and D12) stored in the database 30. However, it may be possible that a part of the data stored in the database 30 is excluded from the input data IN for prediction.

[0069] 13 shows an operation flowchart of the service providing device 20. In the service providing device 20, first, a data acquisition process of step S21 is executed. In the data acquisition process of step S21, the data acquisition unit F1 acquires data D1 to D6, and the application cooperation unit F4 acquires data D11 and D12. The data acquisition process is executed continuously, and the data acquired by the data acquisition process is sequentially accumulated in the database 30.

[0070] After step S21, in step S22, a prediction command including prediction condition data is received by the service providing device 20. The operator of the service use device 40 inputs the prediction condition data to the service use device 40 and then inputs an operation to the service use device 40 to instruct the execution of people flow prediction processing. In response to the input of the operation, a prediction command including the prediction condition data is transmitted from the service use device 40 to the service providing device 20. After step S22, the process proceeds to step S23.

[0071] In step S23, the controller 21 of the service providing device 20 inputs the input data for prediction IN to the simulator F2 in response to receiving the prediction command. The contents of the input data for prediction IN are as described above. The prediction condition data in the input data for prediction IN in step S23 is the prediction condition data in the prediction command. After step S23, the process proceeds to step S24. In step S24, the simulator F2 performs a simulation based on the input data for prediction IN. The simulation performed by the simulator F2 corresponds to a people flow prediction process. In the people flow prediction process based on the input data for prediction IN, the simulator F2 predicts the people flow (in other words, the wandering behavior of visitors to the area) in the prediction target area during the prediction target period, and generates prediction result data OUT indicating the result of the prediction. Thereafter, in step S25, the result output unit F3 transmits the prediction result data OUT to the service use device 40.

[0072] Referring to FIG. 14, the time when the simulator F2 makes a prediction is referred to as a prediction time t X At the prediction time t Xis the timing at which the above-mentioned people flow prediction process is executed. The prediction period is the prediction time t X It is a future period from the perspective of X (The period after the prediction time t X Based on the data stored in the database 30, the prediction time t X The pedestrian flow prediction process is performed at the prediction time t X The earlier period is referred to as the reference period.

[0073] By using the data in the database 30, it is possible to perform a simulation adapted to detailed conditions. As a result, the simulator F2 can be effectively used to verify the effectiveness of visitor attraction measures that are planned daily by each facility operator (PDCA: Plan, Do, Check, Action). In this case, a member of the facility operator may be the operator of the service utilization device 40.

[0074] By performing a simulation using big data D1 on people flow generated throughout the year as input data IN for prediction, it becomes possible to predict people flow taking seasonal influences into account, enabling more precise predictions of people flow (predictions of each person's movement behavior) than the reference simulator. By performing a simulation by including the weather history data D4 and the weather forecast data D5 in the prediction input data IN, it becomes possible to predict people flow taking into account the effects of weather, making it possible to make more precise predictions of people flow (predictions of each person's movement behavior) than with the reference simulator. By performing a simulation by including the temporary timetable data D3a_2 in the prediction input data IN, it becomes possible to predict people flow taking into account the effects of the temporary timetable, enabling more precise prediction of people flow (prediction of each person's movement behavior) than the reference simulator. By performing a simulation using the event history data D6 in the prediction input data IN, it becomes possible to predict people flow taking into account the impact of events, enabling more precise prediction of people flow (prediction of each person's movement behavior) than with the reference simulator. By performing a simulation by including the application log data D12 in the prediction input data IN, it becomes possible to predict people flow using information with excellent real-time performance, making it possible to make more precise people flow predictions (predictions of each person's movement behavior) than with the reference simulator.

[0075] Below, specific operation examples, related technologies, applied technologies, modified technologies, etc. related to the simulator F2 will be described in multiple examples. The matters described above in this embodiment are applied to each of the following examples unless otherwise specified and unless there is a contradiction. If there are matters in each example that contradict the matters described above, the description in each example may take precedence. Furthermore, unless there is a contradiction, matters described in any of the multiple examples shown below can also be applied to any other example (i.e., any two or more of the multiple examples can also be combined).

[0076] <<First Example>> A first embodiment will be described below. In the first embodiment, a people flow prediction process will be described with particular attention to events.

[0077] 15, in a certain area of ​​interest, a first event is held in a period 631, a second event is held in a period 632, and no events are held in a period 633. For the sake of clarity and specificity, it is assumed that only the first and second events can be held in the area of ​​interest, and that the first and second events are not held simultaneously. Therefore, only the first event is held in the period 631, only the second event is held in the period 632, and neither the first event nor the second event is held in the period 633.

[0078] The first event and the second event are two types of events with different contents that can be held in the attention area. For example, if there is a sports facility in the attention area, a track and field sports tournament is held at the sports facility as the first event, and a professional soccer match is held at the sports facility as the second event. Or, for example, if there is a shopping facility in the attention area, a show for children is held at the shopping facility as the first event, and a lottery is held at the shopping facility as the second event.

[0079] Periods 631 to 633 are all forecasted at time t X 15, period 632 is after period 631 and before period 633, but the order of periods 631 to 633 is arbitrary. Except for the fact that an event is held or not and the contents of the event are different, the weather, season, day of the week, and transportation operation conditions are the same for periods 631 to 633. In FIG. 15, period 640 is a period from the prediction time t X In the prediction condition data (see FIG. 12(b)), the attention area is set as the prediction target area, and the period 640 is set as the prediction target period.

[0080] In the first embodiment, at the prediction time t X The event history data D6 in the period 631 (see FIG. 6) includes data indicating whether an event was held in the prediction target area in the period 631 to 633 and the results of the contents of the event. That is, the event history data D6 includes event result data indicating that a first event was held in the prediction target area in the period 631 and event result data indicating that a second event was held in the prediction target area in the period 632. The event result data associated with the period 633 is not included in the event history data D6. For this reason, it is considered that no events were held in the prediction target area in the period 633.

[0081] Regarding the setting of the additional condition data, three cases CS1_1 to CS1_3 are considered. The prediction condition data includes event schedule data as additional condition data (see FIG. 12(b)). The event schedule data indicates whether or not an event will be held in the prediction target period and prediction target area, and the schedule of the event.

[0082] In case CS1_1, first event schedule data indicating a schedule to hold a first event in a prediction target period and a prediction target area is set as additional condition data. In case CS1_2, second event schedule data indicating a schedule to hold a second event in a prediction target period and a prediction target area is set as additional condition data. In case CS1_3, third event schedule data indicating a schedule that no events will be held in the prediction target period and prediction target area is set as the additional condition data.

[0083] In case CS1_1, the people flow prediction process is executed under the condition that a first event is carried out in the prediction target period and prediction target area. In case CS1_2, the people flow prediction process is executed under the condition that a second event is carried out in the prediction target period and prediction target area. In case CS1_3, the people flow prediction process is executed under the condition that no events are held during the prediction period and in the prediction area.

[0084] In the controller 21, the prediction time t XBased on the stored data of the database 30 (hence the prediction input data IN) in the period 631, 632, and 633, the impact of the implementation of the first and second events on the pedestrian flow in the area of ​​interest can be modeled. That is, the pedestrian flow in the area of ​​interest in the periods 631, 632, and 633 can be specified from the pedestrian flow big data D1 in the periods 631, 632, and 633. The pedestrian flow big data D1 in the period 631 refers to the pedestrian flow big data D1 including the movement history data of each user in the period 631. The pedestrian flow big data D1 in the periods 632 and 633 can be interpreted in the same way. Then, the impact of the implementation of the first and second events on the pedestrian flow in the area of ​​interest can be modeled from a comparison between the pedestrian flow in the area of ​​interest in the period 631, the pedestrian flow in the area of ​​interest in the period 632, and the pedestrian flow in the area of ​​interest in the period 633. The algorithm based on this modeling is reflected in the simulator F2. Therefore, by setting event schedule data in the people flow prediction process, it becomes possible to make an accurate people flow prediction based on whether or not an event will be held in the area of ​​interest and the scheduled content of the event.

[0085] By comparing the results of the people flow prediction process for cases CS1_1 to CS_3, the operator of the service use device 40 can know in detail the visitor attraction effect of the event. As a simple example, consider a case where the number of visitors to the area of ​​interest in period 631 is 1.5 times and the number of visitors to the area of ​​interest in period 632 is 2.0 times as many as the number of visitors to the area of ​​interest in period 633. In this case, the predicted number of visitors to the area of ​​interest (i.e., the prediction target area) in period 640 is 1.5 times in case CS1_1 and 2.0 times in case CS1_2 as compared to case CS1_3.

[0086] <<Second Example>> A second embodiment will now be described. As in the first embodiment, in the second embodiment, people flow prediction processing will be described with particular attention to events.

[0087] Fig. 16 shows an area of ​​interest 660 that is set as a prediction target area. Within the area of ​​interest 660, there are one or more facilities including a shopping facility 661 as a target facility (only the shopping facility 661 is shown in Fig. 16).

[0088] 17, it is assumed that a specific event is held in a period 672 in an area of ​​interest 660, and that no events, including the specific event, are held in periods 671 and 673. X 17, period 672 is a period after period 671, but period 672 may be a period before period 671. Period 673 is a period after periods 671 and 672.

[0089] The specific event is an event that can be held at the shopping facility 661. Business of the shopping facility 661 without a specific event is called normal business. For example, the specific event is a discount event in which products are sold at a discount compared to normal business. The specific event may be an event that involves the issuance of a coupon that can be used at the shopping facility 661. The coupon for the specific event can be used at the shopping facility 661 only during the period in which the specific event is held. The specific event includes the issuance of a coupon that can be used at the shopping facility 661, and may further include other events. Alternatively, for example, the specific event is a show event in which a show aimed at a certain age group is held. In addition to the above, the specific event may take any form.

[0090] Except for the difference in whether or not a specific event is held and the difference in the business environment described later, the weather, season, day of the week, and operation status of transportation are the same between periods 671 and 673. In FIG. 17, the period 680 is the prediction time t XIn the prediction condition data (see FIG. 12(b)), an area of ​​interest 660 is set as the prediction target area, and a period 680 is set as the prediction target period. Hereinafter, the code "660" may be used as the code for the prediction target area, and the code "680" may be used as the code for the prediction target period.

[0091] In the second embodiment, at the prediction time t X The event history data D6 in the period 672 (see FIG. 6) includes data indicating whether an event was held in the prediction target area 660 in the periods 671 to 673 and the results of the contents of the event. That is, the event history data D6 includes event result data indicating that a specific event was held in the prediction target area 660 in the period 672. The event result data associated with the period 671 and the event result data associated with the period 673 are not included in the event history data D6. For this reason, it is considered that no events were held in the prediction target area 660 in the periods 671 and 673.

[0092] Here, the periods 671 to 673 and 680 are respectively periods from 11:00 to 18:00 set within one day. A day including the period 671 is called the first date. A day including the period 672 is called the second date. A day including the period 673 is called the third date.

[0093] Figure 18(a) shows visitor performance data 671D of the shopping facility 661 on the first date. Figure 18(b) shows visitor performance data 672D of the shopping facility 661 on the second date. Figure 18(c) shows visitor performance data 673D of the shopping facility 661 on the third date.

[0094] Visitor record data 671D is record data on the number of visitors to shopping facility 661 on the first date, and indicates the total number of people who visited shopping facility 661 on the first date in hourly units. Furthermore, visitor record data 671D also indicates the attribute distribution of people who visited shopping facility 661 on the first date. In Fig. 18(a), the sizes of the black area, dotted area, and diagonal line area respectively indicate the numbers of people with the first, second, and third attributes who visited shopping facility 661 on the first date. Visitor record data 672D is record data on the number of visitors to shopping facility 661 on the second date, and indicates the total number of people who visited shopping facility 661 on the second date in hourly units. Furthermore, visitor record data 672D also indicates the attribute distribution of people who visited shopping facility 661 on the second date. In Fig. 18(b), the sizes of the black area, dotted area, and diagonal line area respectively indicate the number of people with the first, second, and third attributes who visited shopping facility 661 on the second date. Visitor record data 673D is record data on the number of visitors to shopping facility 661 on the third date, and indicates the total number of people who visited shopping facility 661 on the third date in hourly units. Furthermore, visitor record data 673D also indicates the attribute distribution of people who visited shopping facility 661 on the third date. In Fig. 18(c), the sizes of the black area, dotted area, and diagonal line area respectively indicate the number of people with the first, second, and third attributes who visited shopping facility 661 on the third date.

[0095] The first to third attributes are three different types of attributes. For example, a person with the first attribute is a man in his 20s or older, a person with the second attribute is a woman in his 20s or older, and a person with the third attribute is a man and a woman under his 20s. Four or more types of attribute distributions may be shown in the visitor record data 671D to 673D (the same applies to visitor prediction data 680D described later).

[0096] The controller 21 (e.g., simulator F2) can generate visitor performance data 671D from the people flow big data D1 on the first date (people flow big data D1 including movement history data of each user on the first date). The controller 21 (e.g., simulator F2) can generate visitor performance data 672D from the people flow big data D1 on the second date (people flow big data D1 including movement history data of each user on the second date). The controller 21 (e.g., simulator F2) can generate visitor performance data 673D from the people flow big data D1 on the third date (people flow big data D1 including movement history data of each user on the third date).

[0097] Now, consider that the commercial environment of the attention area 660 has changed at a certain boundary period. For example, assume that the number of visitors to the shopping facility 661 has tended to decrease after the boundary period compared to before the boundary period due to the influence of an epidemic (such as an epidemic caused by the new coronavirus) that occurred at the boundary period. Periods 671 and 672 are periods before the boundary period (for example, periods belonging to 2019), and periods 673 and 680 are periods after the boundary period (for example, periods belonging to 2023). The decrease in the number of visitors due to the change in the commercial environment can be seen from a comparison of visitor performance data 671D and 673D.

[0098] After the third date, which includes the period 673, the operator of the shopping facility 661 can use the simulator F2 to determine how many visitors can be expected when a specific event is held again in the prediction period 680.

[0099] FIG. 18(d) shows visitor prediction data 680D of the shopping facility 661 on a date including the period 680. A day including the period 680 is called a predicted date. The visitor prediction data 680D is predicted data of the number of visitors to the shopping facility 661 on the predicted date, and shows a predicted value of the total number of people visiting the shopping facility 661 on the predicted date in one-hour units. Furthermore, the visitor prediction data 680D also shows a prediction of the attribute distribution of people visiting the shopping facility 661 on the predicted date. In FIG. 18(d), the sizes of the blackened area, dotted area, and hatched area respectively represent the predicted numbers of people with the first, second, and third attributes visiting the shopping facility 661 on the predicted date.

[0100] The prediction condition data includes event schedule data as additional condition data (see FIG. 12(b)). The event schedule data in the example of FIG. 18(d) represents a schedule for holding a specific event in a shopping facility 661 during a prediction period 680. The simulator F2 executes a people flow prediction process under the conditions that the specific event will be held in the shopping facility 661 during the prediction period 680. At this time, the simulator F2 generates visitor prediction data 680D based on the visitor actual data 671D to 673D, and includes the visitor prediction data 680D in the prediction result data OUT.

[0101] The simulator F2 can derive an increase rate of the number of visitors accompanying the implementation of a specific event based on the visitor actual data 671D and 672D. The simulator F2 can generate visitor prediction data 680D based on the derived increase rate and the visitor actual data 673D. That is, for example, it is possible to estimate that the number of visitors will increase by the above-mentioned increase rate on the predicted date based on the number of visitors shown in the visitor actual data 673D, and to show the estimated result in the visitor prediction data 680D. Commercial environment information that the commercial environment has changed as described above between the period to which the first date and the second date belong and the period to which the third date and the predicted date belong may be provided to the simulator F2, and the people flow prediction process may be performed taking into account the commercial environment information.

[0102] The method shown in this embodiment allows for accurate prediction of future customer attraction effects of holding an event by taking into consideration past visitor performance data (671D and 672D) in a different business environment from the present and current visitor performance data (673D). By improving the prediction accuracy, the investment effect of holding the event can be precisely verified.

[0103] The operator of the shopping facility 661 can consider and decide whether or not to hold a particular event based on the visitor prediction data 680D, or can consider whether or not to hold a different, better event.

[0104] When the operator of the shopping facility 661, having checked the visitor prediction data 680D, decides to hold a specific event, the operator may input a specific operation indicating the decision to the service use device 40. In response to the input of the specific operation, specific information is transmitted from the service use device 40 to the service providing device 20. In response to receiving the specific information, the application cooperation unit F4 may transmit, to each terminal device 10 having the travel assistance application installed, travel guidance information indicating that a specific event will be held in the shopping facility 661. However, the terminal device 10 to which the travel guidance information is transmitted may be limited in consideration of the distance between the shopping facility 661 and the terminal device 10. If the specific event involves the issuance of a coupon that can be used in the shopping facility 661, the coupon may be added to the travel guidance information. The same applies when it is decided to hold another event.

[0105] <<Third Example>> A third embodiment will be described. In the third embodiment, an applied technology to the second embodiment will be described. As described above, terminal location information of the terminal device 10[i] on which the migration support app is installed is transmitted in real time to the service providing device 20 as location data of the user U[i], and is reflected in the application log data D12. Hereinafter, the terminal device 10 on which the migration support app is installed may be referred to as a specific terminal device 10, and the user of the specific terminal device 10 may be referred to as a specific user.

[0106] The service providing device 20 can obtain location data of a plurality of specific users from the corresponding specific terminal devices 10 in real time, and can predict the location data at the prediction time t X The application log data D12 includes the location data at the prediction time t X This makes it possible to predict visitors with high accuracy according to the distribution of people within and around the prediction target area 660.

[0107] We will explain this in more detail. In order to improve the accuracy of prediction, X It is better to be closer to the forecast period 680. For the sake of concrete explanation, X is a point in time prior to the forecast target period 680, but is within the forecast date to which the forecast target period 680 belongs.

[0108] Simulator F2 predicts the time t X Based on the application log data D12 at the prediction time t X Based on the identification result, the simulator F2 adjusts the predicted value of the total number of people visiting the shopping facility 661 on the predicted date. In the third embodiment, the adjusted predicted value is shown in the visitor prediction data 680D.

[0109] That is, for example, the simulator F2 predicts the time t X In the process, the number of specific users NUM located within a specified radius of the shopping facility 661 is counted. For example, the simulator F2 increases the predicted value of the total number of people visiting the shopping facility 661 on the predicted date as the number of people NUM increases. In this case, after obtaining the predicted value of the total number of people visiting the shopping facility 661 on the predicted date by the method shown in the second embodiment, the predicted value may be adjusted according to the number of people NUM. Simply put, for example, in the adjustment, the predicted value may be adjusted by adding (NUM / NUM REF ) multiplied by (NUM REF is a standard number set based on the total number of specific users, etc.)

[0110] The predicted value may be adjusted for each attribute of a specific user. That is, the simulator F2 adjusts the predicted value for the prediction time t X In the example, the simulator F2 counts the number NUM[j] of specific users with the jth attribute who are located within a predetermined radius from the shopping facility 661. For example, the simulator F2 increases the predicted value for the total number of people with the jth attribute who will visit the shopping facility 661 on the predicted date as the number NUM[j] increases. Here, j represents 1, 2, or 3.

[0111] The method shown in this embodiment is particularly useful in an environment where the mobility support application is widely used. The operator of the shopping facility 661 can consider and decide whether to hold a specific event based on the visitor prediction data 680D, or can consider whether to hold a better event.

[0112] When the operator of the shopping facility 661, having checked the visitor prediction data 680D, decides to hold a specific event, the operator may input a specific operation indicating the decision to the service use device 40. In response to the input of the specific operation, specific information is transmitted from the service use device 40 to the service providing device 20. In response to receiving the specific information, the application cooperation unit F4 may transmit, to each specific terminal device 10, tour guide information indicating that a specific event will be held in the shopping facility 661. However, the specific terminal device 10 to which the tour guide information is transmitted may be limited in consideration of the distance between the shopping facility 661 and the specific terminal device 10. If the specific event involves the issuance of a coupon that can be used in the shopping facility 661, the coupon may be added to the tour guide information. The same applies when it is decided to hold another event.

[0113] <<Fourth Example>> A fourth embodiment will now be described. As already mentioned, the input data IN for prediction includes weather history data D4 and weather forecast data D5. When focusing on the prediction target area and prediction target period, the weather history data D4 includes the weather history of the prediction target area (i.e., represents the weather history of the prediction target area). The weather forecast data D5 includes a forecast of the weather of the prediction target area during the prediction target period (i.e., represents the weather forecast of the prediction target area).

[0114] For this reason, simulator F2 is capable of predicting people flow taking into account the effects of weather, enabling more precise prediction of people flow (prediction of each person's movement behavior) than the reference simulator.

[0115] The impact of past weather on the movement behavior of visitors to the prediction target area can be modeled from the people flow big data D1 for the reference period and the weather history of the prediction target area for the reference period. Based on the modeling results and the weather forecast for the prediction target area for the prediction target period, the simulator F2 can precisely predict the people flow in the prediction target area for the prediction target period.

[0116] <<Fifth Example>> A fifth embodiment will be described. As already mentioned, the prediction input data IN includes train schedule data D3a. The train schedule data D3a includes at least basic train schedule data D3a_1, and when a special operation period is set, includes special train schedule data D3a_2 that is valid only during the special operation period. The special train schedule data D3a_2 is set temporarily after the basic train schedule data D3a_1 is set.

[0117] For this reason, simulator F2 is capable of predicting people flow taking into account the effects of special timetables, and is capable of more precise prediction of people flow (prediction of each person's movement behavior) than the reference simulator.

[0118] Consider the case where temporary schedule data D3a_2 is set for the prediction period. That is, consider the case where the prediction period overlaps with a certain temporary operation period. From the people flow big data D1 for the reference period and the operation schedule data D3a for the reference period, it is possible to model the impact that the temporary schedule had on the movement behavior of visitors to the prediction target area in the past. Based on the modeling results and the temporary schedule data D3a_2 applied to the prediction period, the simulator F2 can precisely predict the people flow within the prediction target area in the prediction period.

[0119] In addition, when the special operation period is set in a planned manner, the special operation period is set to a time point (for example, one week or one month) prior to the start time of the special operation period. When the prediction target period overlaps with a certain special operation period, the prediction time t X Assume that the special schedule data D3a_2 corresponding to a specific special operation period has already been stored in the database 30. Therefore, at the prediction time t X In this case, it becomes possible to predict people flow taking into account the temporary schedule data D3a_2 applied to the prediction period.

[0120] A temporary schedule may be suddenly set due to a sudden traffic disturbance caused by bad weather or an accident (therefore, a temporary operation period may be suddenly set). Even in this case, the temporary schedule data D3a_2 regarding the temporary schedule that is suddenly set is stored at the prediction time t X It is assumed that the data has already been stored in the database 30.

[0121] <<Sixth Example>> A sixth embodiment will be described. As already mentioned, the service providing device 20 can obtain location data of a plurality of specific users from the corresponding specific terminal devices 10 in real time. Then, the service providing device 20 obtains the location data of the plurality of specific users from the corresponding specific terminal devices 10 in real time at the prediction time t X The application log data D12 including the location data of each specific user in the above can be included in the prediction input data IN. The specific terminal device 10 refers to the terminal device 10 in which the travel assistance application is installed, and the specific user refers to the user of the specific terminal device 10.

[0122] By performing a simulation using the application log data D12 as input data IN for prediction, it is possible to predict people flow using information with excellent real-time characteristics, and it is possible to predict people flow (prediction of each person's movement behavior) more precisely than with the reference simulator. X This becomes evident when the prediction period is close enough to require real-time performance, and when a larger number of people are specific users. In the extreme case where all people in the prediction area are specific users, the simulator F2 can accurately grasp the current flow of people in the prediction area in real time, and can therefore predict the flow of people in the prediction area in the near future with high accuracy.

[0123] When the ratio of specific users to all users is sufficiently high, the application registration data D11 and the application log data D12 may be used as people flow big data with excellent real-time characteristics instead of the people flow big data D1 from the telecommunications carrier.

[0124] <<Seventh Example>> A seventh embodiment will be described. Up to this point, it is assumed that a business operator or the like uses the people flow prediction service, and the method for setting the prediction target area and the prediction target period according to the input operation contents to the service use device 40 has been described. However, the method for setting the prediction target area and the prediction target period is not limited to this. Focusing on a terminal device 10[i] on which a migration support app is installed, the migration support service provided to a user U[i] will be described.

[0125] The simulator F2 may set the prediction target area based on the terminal location information of the terminal device 10[i] (hence the location data of the user U[i]). In this case, the simulator F2 may set the prediction target area to an area including the current location of the user U[i] based on the current location of the user U[i] (the location of the user U[i] at the current time). The simulator F2 may set the prediction target period to a period having a certain length of time in the near future as viewed from the present time. After setting the prediction target area and the prediction target period, the simulator F2 may perform a people flow prediction process based on the prediction input data IN, and transmit the obtained prediction result data OUT to the terminal device 10[i].

[0126] The simulation in the people flow prediction process is performed taking into consideration the latest traffic information (operation-related data D3) available at the present time. Therefore, it is possible to predict the movement behavior of visitors to the prediction target area, including the utilization status of the transportation facilities entering the prediction target area. By linking the results of the simulation to the movement support app, it is possible to add a traffic guidance function that avoids congestion concentration to the movement support app. That is, in the terminal device 10[i], the movement support app can suggest a destination to the user U[i] and a means of transportation to the destination based on the latest traffic information (the latest operation-related data D3 or data equivalent thereto). In suggesting a means of transportation, the terminal device 10[i] can make an optimal proposal in accordance with the movement support app so as to avoid congestion concentration and taking into consideration the setting status of a temporary timetable.

[0127] Transportation schedules may be disrupted due to sudden traffic disruptions caused by bad weather or accidents, etc. Even in such cases, by conducting simulations that take temporary schedules into account, it is possible to provide travel support services, including appropriate transportation guidance, to user U[i].

[0128] <<Eighth Example>> An eighth embodiment will be described. A discrete choice model can be used in the simulation of people flow prediction processing. An example of the discrete choice model will be described.

[0129] For convenience of explanation, consider a situation in which any of points a, b, and c can be the destination of a model person. A model person is a person assumed in a discrete choice model (a modeled virtual person), and can be thought of as imitating an arbitrary user U[i].

[0130] In this case, the utility functions Fa, Fb, and Fc can be expressed by the following formulas (1) to (3). Here, it is assumed that the utility felt by the model person for each of points a, b, and c can be expressed by seven explanatory variables. The seven explanatory variables are "age," "gender," "occupation," "purpose," "weather," "season," and "event." Fa=a1+a2+a3+a4+a5+a6+a7 ···(1) Fb=b1+b2+b3+b4+b5+b6+b7 ···(2) Fc=c1+c2+c3+c4+c5+c6+c7...(3)

[0131] The explanatory variables "age," "gender," and "occupation" respectively represent the age, gender, and occupation of the model person. The explanatory variable "purpose" represents the purpose for which the model person travels to the area including points a to c, such as shopping, work, dining, leisure, etc. The explanatory variable "weather" represents the weather in the prediction target area including points a to c. The explanatory variable "season" represents the season in the prediction target area including points a to c. In this embodiment, since the prediction target area including points a to c is an area within Japan, the explanatory variable "season" represents the season of the prediction target period in Japan. The explanatory variable "event" represents whether or not an event is held in the prediction target area including points a to c, and the content of the event.

[0132] The values ​​of the utility functions Fa, Fb, and Fc represent the utility that the model person obtains by going to points a, b, and c, respectively.

[0133] a1 to a7 are coefficients for expressing the utility that point a provides to the model person for the explanatory variables "age", "sex", "occupation", "purpose", "weather", "season", and "event", respectively, and are defined as functions. b1 to b7 are coefficients for expressing the utility that point b provides to the model person for the explanatory variables "age", "sex", "occupation", "purpose", "weather", "season", and "event", respectively, and are defined as functions. c1 to c7 are coefficients for expressing the utility that point c provides to the model person for the explanatory variables "age", "sex", "occupation", "purpose", "weather", "season", and "event", respectively, and are defined as functions.

[0134] The functions a1, b1, and c1 are functions that use the age of the model person as a variable, and have values ​​that depend on the age of the model person. Functions a2, b2, and c2 are functions that use the gender of the model person as a variable, and have values ​​that depend on the gender of the model person. Functions a3, b3, and c3 are functions that use the occupation of the model person as a variable, and have values ​​that depend on the occupation of the model person. Functions a4, b4, and c4 are functions that use the model person's purpose as a variable, and have values ​​that depend on the model person's purpose. Functions a5, b5, and c5 are functions that use the weather in the prediction target area including points a to c as a variable, and have values ​​that depend on the weather. Functions a6, b6, and c6 are functions that use the season of the prediction target area including points a to c as a variable, and have values ​​that depend on the season. Functions a7, b7, and c7 are functions whose variables are the implementation status of the event in the prediction target area including points a to c, and have values ​​that depend on whether the event is implemented and the implementation details.

[0135] In a discrete choice model, prediction time t X Based on the data previously stored in the database 30, functions a1 to a7, b1 to b7 and c1 to c7 are defined by known multivariate analysis (multinomial logistic regression analysis).

[0136] After defining the functions a1 to a7, b1 to b7, and c1 to c7, a simulation of the people flow prediction process is performed. In the simulation, it is assumed that for each visitor to the prediction target area, the visitor selects the point from points a, b, and c that provides the highest utility as the destination, and the destinations of all visitors are predicted. Based on this prediction result, prediction result data OUT is generated.

[0137] In the reference simulator mentioned above, the utility function is defined using only four explanatory variables "age," "gender," "occupation," and "purpose," based on person trip data collected over a short period of time once every few years. This makes it impossible to faithfully reproduce the tendency of visitors to select a destination, which is affected by short-term variable factors such as weather, season, and the presence or absence of events.

[0138] In response to this, simulator F2 performs a simulation (people flow prediction process) using utility functions Fa to Fc according to formulas (1) to (3) that also take into account "weather," "season," "events," etc. Therefore, it is possible to predict the tendency of visitors to select destinations, taking into account short-term variable factors such as weather, season, and the presence or absence of events. Facility operators who wish to attract customers can use simulator F2 to predict the effects of daily attraction measures and develop better implementation plans for the measures.

[0139] Although the example given here is a definition of a model for selecting a destination, a model for selecting a means of transportation can be defined in a similar manner. Therefore, the method shown in this embodiment can also be applied to a transportation guidance system based on transportation demand.

[0140] <<Ninth Example>> A ninth embodiment will be described. Fig. 19 shows a flow of processing that can be executed by the simulator F2. The processing of steps S61 to S66 shown in Fig. 19 can be incorporated into the people flow prediction processing of the simulator F2, and can therefore be performed in step S24 of Fig. 13. Input data for prediction IN is input to the simulator F2 (step S23 of Fig. 13), and the simulator F2 estimates visitors to the prediction target area in the prediction target period based on the input data for prediction IN. In step S61, attributes of each visitor are set. The attributes of each visitor are set based on the people flow big data D1.

[0141] In the next step S62, the simulator F2 estimates whether each visitor to the prediction target area will continue their wandering behavior within the prediction target area or return home. In the next step S63, the simulator F2 estimates the destination of a visitor who is estimated to continue their wandering behavior within the prediction target area (hereinafter referred to as a continuing wandering visitor). In the next step S64, the simulator F2 estimates the means of transportation that the continuing wandering visitor will use to travel to the destination. In the next step S65, the simulator F2 estimates the stay time of the continuing wandering visitor at the destination.

[0142] Thereafter, in step S66, the simulator F2 generates output data based on the estimation results of steps S62 to S65. For example, the output data generated in step S66 includes a predicted value of the stay time of each visitor in the prediction target area and the travel distance of each visitor in the prediction target area. Also, for example, the output data generated in step S66 includes a transition in the number of visitors to each store in the prediction target area. Also, for example, the output data generated in step S66 includes the frequency of use of transportation facilities by each visitor to the prediction target area. The output data generated in step S66 may be included in the prediction result data OUT (see step S25 in FIG. 13).

[0143] The estimation in steps S62 to S65 is performed based on the input data IN for prediction. The input data IN for prediction includes data related to weather, special schedules, and events, making it possible to make a precise estimation that takes these data into consideration. If additional condition data related to incentive measures is added to the input data IN for prediction (adding event schedule data, etc.: FIG. 12(b)), the addition will affect the estimation of destinations, etc. Based on the prediction result data OUT that takes into account this effect, businesses considering incentive measures can verify the effectiveness of the incentive measures.

[0144] <<Additional Notes>> An addendum is provided for each of the above items.

[0145] The simulator F2 executes a people flow prediction process using a people flow prediction method. The people flow prediction method is a method for predicting people flow in a prediction target area during a prediction target period based on prediction input data IN (a method for predicting the movement behavior of each person in the prediction target area). The prediction target period is the time point at which prediction is performed (prediction time t X ) is a period after the prediction target area. The prediction input data IN includes human flow big data D1 including attribute data of each user and movement history data of each user. The users are people whose attribute data and movement history data are collected in the human flow prediction method, and there are many of them. The prediction input data IN includes event history data D6, which indicates whether an event is held in the prediction target area before the prediction target period and the actual contents of the event. In this case, the prediction input data IN may further include event schedule data (see FIG. 12(b)). The event schedule data indicates whether an event is held in the prediction target area during the prediction target period and the schedule of the contents of the event. The event in the prediction target area may be the issuance of a coupon that can be used at any facility located in the prediction target area, and may further include other events other than the issuance of the coupon. The coupon for the event in the prediction target area can be used only during the implementation period of the event.

[0146] From the people flow big data D1 and event history data, it is possible to obtain actual data on visitors (visitors to the prediction target area) when an event is held and actual data on visitors when no event is held. It is also possible to know how many visitors there were when what kind of event was held. Therefore, if event schedule data is included in the prediction input data IN, it is possible to perform highly accurate people flow prediction according to the prediction target period and whether or not an event will be held in the prediction target area and the schedule of the event's contents. In other words, it is possible to perform appropriate people flow prediction taking into account short-term variable factors due to events (variable factors that affect people flow).

[0147] The service providing device 20 includes the functionality of a people flow prediction device, and when attention is focused only on the people flow prediction process, the service providing device 20 functions as a people flow prediction device.

[0148] A program for causing a computer device to execute any of the methods described in the embodiments of the present invention, and a non-volatile recording medium having the program recorded thereon, are included in the scope of the embodiments of the present invention. The program for causing a computer device to execute any of the methods described in the embodiments of the present invention may be a subprogram incorporated into any main program or called by any main program. The people flow prediction program is a program for causing a service providing device 20 (controller 21) as a computer device to execute the people flow prediction method. Any of the processes in the embodiments of the present invention may be realized by hardware such as a semiconductor integrated circuit, software equivalent to the above program, or a combination of hardware and software.

[0149] The embodiments of the present invention can be modified in various ways as appropriate within the scope of the technical ideas set forth in the claims. The above embodiments are merely examples of the present invention, and the meanings of the terms of the present invention and each component are not limited to those described in the above embodiments. The specific numerical values ​​shown in the above description are merely examples, and can be changed to various numerical values ​​as a matter of course. [Explanation of symbols]

[0150] SYS system NET Communication network 10 Terminal Equipment 20 Service provision equipment 21 Controller 30 Databases 40 Service utilization device 50 Business server equipment F1 Data Acquisition Department F2 Simulator F3 Result output section F4 App Integration Department D1 People Flow Big Data D2 Map Data D3 Operation-related data D3a Timetable data D3a_1 Basic Diagram Data D3a_2 Special timetable data D3b Congestion data D4 Weather History Data D5 Weather Forecast Data D6 Event History Data D11 App registration data D12 App log data

Claims

1. A people flow prediction method for predicting people flow in a prediction target area during a prediction target period based on input data, The input data is People flow big data including attribute data of each user and movement history data of each user; Event history data indicating whether or not an event was held in the prediction target area prior to the prediction target period and the details of the event; and event schedule data indicating whether or not an event will be held and the details of the event in the prediction target period and the prediction target area. , a method for predicting the flow of people.

2. The target facility is located within the prediction target area, Acquire, as first performance data, performance data on the number of visitors to the target facility during a first reference period and when a specific event is not held at the target facility, based on the people flow big data; Acquire, as second performance data, performance data on the number of visitors to the target facility during a second reference period and when a specific event is held at the target facility, based on the people flow big data; Acquire, as third performance data, performance data on the number of visitors to the target facility during a third reference period and when the specific event is not held at the target facility, based on the people flow big data; the first reference period and the second reference period are periods prior to the prediction period, and the third reference period is a period subsequent to the first reference period and the second reference period and prior to the prediction period; generating prediction data for the number of visitors to the target facility when the specific event is held at the target facility during the prediction period based on the first performance data, the second performance data, and the third performance data; The method for predicting people flow according to claim 1 .

3. acquiring location data of a plurality of specific users each having a specific terminal device from the specific terminal device in real time, and including the location data of each specific user in the input data; The location data of each specific user after the third reference period and before the prediction target period is reflected in the prediction data. The method for predicting people flow according to claim 2.

4. The input data is Weather history data including a weather history of the prediction target area; and weather forecast data including a forecast of the weather in the prediction target area during the prediction target period. The method for predicting people flow according to any one of claims 1 to 3.

5. The input data includes transportation schedule data for the prediction target area, The train schedule data includes basic train schedule data and temporary train schedule data that is temporarily set after the basic train schedule data is set. The method for predicting people flow according to any one of claims 1 to 3.

6. Among the users, location data of a plurality of specific users each having a specific terminal device is acquired in real time from the specific terminal device, and the location data of each specific user is included in the input data. The method for predicting people flow according to claim 1 or 2.

7. The event in the prediction target area includes issuing a coupon that can be used at a facility in the prediction target area. The method for predicting people flow according to claim 1 .

8. The specific event at the target facility includes issuing a coupon that can be used at the target facility. The method for predicting people flow according to claim 1 or 2.

9. A people flow prediction device that predicts people flow in a prediction target area during a prediction target period based on input data, The input data is People flow big data including attribute data of each user and movement history data of each user; Event history data indicating whether or not an event was held in the prediction target area prior to the prediction target period and the details of the event; and event schedule data indicating whether or not an event will be held and the details of the event in the prediction target period and the prediction target area. , a human flow prediction device.

10. A people flow prediction program for causing a computer to execute a people flow prediction method for predicting people flow in a prediction target area during a prediction target period based on input data, The input data is People flow big data including attribute data of each user and movement history data of each user; Event history data indicating whether or not an event was held in the prediction target area prior to the prediction target period and the details of the event; and event schedule data indicating whether or not an event will be held and the details of the event in the prediction target period and the prediction target area. , a people flow prediction program.

Images