Return area prediction device

Abstract

A return area prediction device (10) comprises a prediction unit (12) that uses a regularized regression method to obtain a regression result for predicting, from event information about an event of interest, the number of visitors in each return area for the event of interest, identifies, on the basis of the acquired regression result, a nearest-neighbor cluster closest to the regression result among event group clusters that have already been clustered, and predicts the number of people in each return area for the event of interest on the basis of at least the position of the center of gravity of the nearest-neighbor cluster.

Description

[Technical field]

[0001] The present disclosure relates to a return area prediction device that predicts the return area of ​​visitors to various events, activities, performances, etc. (hereinafter collectively referred to as "events"). The "return area" here means the first stay area of ​​the visitor's movement immediately after the event. For example, if the visitor goes straight home immediately after the event, the return area is the area around the home, and if the visitor stops at a commercial facility immediately after the event, the return area is the area around the commercial facility. In addition, "transportation routes" related to the prediction of the return area include all routes that operate along a predetermined route and stop at predetermined stops, such as train routes and bus routes. [Background technology]

[0002] When planning various events, it is important to predict the return area of ​​visitors expected to attend the event in advance for the purpose of considering measures for traffic improvement to the event venue, marketing of goods sales, etc. However, in the past, analysis results obtained by grasping and analyzing attribute information of visitors of past events from ticket sales data of the events, etc., and know-how of event promoters, etc., were used as substitutes for return area information regarding visitors of the planned event to consider measures for traffic improvement to the event venue, marketing of goods sales, etc. In addition, Patent Document 1 discloses a technology that calculates the increase rate of the number of people getting off at the nearest station before the start of the event on the day of the event compared to the number of people getting off at the same station on a day when the event is not held, and predicts the number of users at the nearest station immediately after the event ends using the obtained rate, etc. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2015-219673 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, even if analysis results obtained based on attribute information of attendees of past events or the know-how of event promoters are utilized, or the number of users at the nearest station immediately after the event ends based only on the number of passengers getting off at the nearest station before the event starts on the day of the event, there is a limit to how much the prediction accuracy of the return area of ​​attendees of upcoming events can be improved, and it has been particularly difficult to apply this to predictions of new events or new event venues. Also, since the number of event performances itself is far from the amount of data required for general machine learning, if a machine learning method is applied as is, it will be too faithful to the small amount of training data, resulting in over-learning that deviates from the original data trends, and there is a risk that this will hinder the improvement of the prediction accuracy of the return area.

[0005] The present disclosure has been made to solve the above-mentioned problems, and aims to improve the accuracy of predicting the return area of ​​visitors to an upcoming event. [Means for solving the problem]

[0006] The return area prediction device according to the present disclosure includes a prediction unit that uses a regularized regression method to obtain a regression result of predicting the number of attendees of a target event for each return area from event information of the target event, identifies a nearest cluster that is closest to the regression result among event group clusters that have already been clustered based on the obtained regression result, and predicts the number of attendees for the target event for each return area based on at least the center of gravity of the nearest cluster.

[0007] In the above-mentioned return area prediction device, the prediction unit uses a regularized regression method to obtain a regression result of the prediction of the number of visitors to each return area of ​​the target event from the event information of the target event, and identifies the nearest cluster to the regression result from among the already clustered event group clusters based on the obtained regression result. The prediction unit then predicts the number of visitors to each return area of ​​the target event based at least on the center of gravity of the identified nearest cluster. In this way, the nearest cluster that is closest to the regression result by the regularized regression method is identified, and the number of visitors to each return area of ​​the target event is predicted based at least on the center of gravity of the identified nearest cluster. This makes it possible to accurately predict the number of visitors to each return area of ​​the target event even if the event has never been held in the past or has only a small history of being held, or even if tickets have not yet been released.

[0008] Furthermore, when predicting the number of people for each return area of ​​the target event, the prediction unit uses at least the center of gravity of the nearest cluster as basic information. In addition to the center of gravity of the nearest cluster, the regression result may be used as the basic information at this time, but using only the regression result is avoided. This prevents overlearning, which is a case where the regression result obtained by regression from the event information of the target event is too faithfully adapted to deviate from the original data trend, and avoids a situation in which the prediction accuracy of the return area is hindered. Effect of the Invention

[0009] According to the present disclosure, it is possible to improve the accuracy of predicting the return area of ​​attendees of a scheduled event. [Brief description of the drawings]

[0010] [Figure 1] 1 is a functional block diagram of a return area prediction device and related devices according to an embodiment of the present invention. [Diagram 2] FIG. 11 is a diagram illustrating an example of location information collected by a location information collection process. [Diagram 3]13 is a diagram showing an example of event information regarding a group of past events acquired at the beginning of the prediction model learning process. FIG. [Figure 4] FIG. 13 is a diagram showing an example of information about event attendees identified in the prediction model learning process. [Diagram 5] (a) is a diagram showing an example of statistical visitor information (by city, ward, town, or village) used to generate a prediction model, (b) is a diagram showing an example of statistical visitor information (by station) used to generate a prediction model, (c) is a diagram showing an example of statistical visitor information (visitors from far away) used to generate a prediction model, and (d) is a diagram showing an example of statistical visitor information (results of location information aggregation before the event starts) used to generate a prediction model. [Figure 6] FIG. 1A is a diagram showing an example of event information for a target event that is acquired at the beginning of the return area prediction process, and FIG. 1B is a diagram showing an example of aggregated location information data before the start of an event that is acquired as an option together with the event information. [Figure 7] FIG. 1(a) is a diagram showing an example of a return area prediction result (by city, ward, town, or village) output by the return area prediction process, FIG. 1(b) is a diagram showing an example of a return area prediction result (by station) output by the return area prediction process, and FIG. 1(c) is a diagram showing an example of a return area prediction result (visitors from far away) output by the return area prediction process. [Figure 8] 13A is a flow diagram showing a location information collection process, FIG. 13B is a flow diagram showing a prediction model learning process, and FIG. 13C is a flow diagram showing a return area prediction process. [Figure 9] (a) is a diagram showing multiple event group clusters obtained by a prediction model learning process, (b) is a diagram showing a process for identifying the nearest cluster, and (c) is a diagram showing a process for predicting the number of people represented by the center of gravity of the nearest cluster as the number of people for each return area of ​​the target event. [Figure 10] FIG. 1A is a flow diagram showing a first modification of a return area prediction process, and FIG. 1B is a diagram for explaining a prediction method in the first modification. [Figure 11]FIG. 13A is a flow diagram showing a second modification of a return area prediction process, and FIG. 13B is a diagram for explaining a prediction method in the second modification. [Figure 12] FIG. 2 is a diagram illustrating an example of a hardware configuration of a return area prediction device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of a return area prediction device according to the present disclosure will be described with reference to the drawings. As shown in Fig. 1, a return area prediction device 10 includes a cluster acquisition unit 11, a prediction unit 12, a location information database (DB) 13, and an event database (DB) 14. Below, the functions of each unit will be outlined.

[0012] The location information DB13 is a database that constantly receives (collects) and stores location information from various user terminals 20. The location information includes, for example, the location of the user (mesh code identifying the mesh in which the user is located) and the information acquisition date and time, as shown in FIG. 2, and these are stored in the location information DB13 together with the user's identifier (user ID).

[0013] The event DB 14 is a database that receives (collects) and stores event information on various past events from the business terminal 30 at any time. The event information includes, for example, the event name, performers, genre, sub-genre, event date, event location, venue opening time, performance start time, and performance closing time, as shown in FIG. 3, and is stored in the event DB 14 together with the event's identifier (ID). The business terminal 30 is a terminal involved in the collection and management of event information, and also inputs event information on a target event to the prediction unit 12, acquires and outputs (displays and outputs) a predicted number of people for each return area for the target event output from the prediction unit 12. However, the latter process (input of event information on a target event, and acquisition and output of a predicted number of people for each return area) may be configured to be performed by a terminal other than the business terminal 30.

[0014] It is not essential that the above-mentioned location information DB13 and event DB14 are provided inside the return area prediction device 10; they may be provided in an external server or the like, and the return area prediction device 10 may be configured to obtain the necessary information from an external server or the like.

[0015] The cluster acquisition unit 11 is a functional unit that calculates the number of visitors to each event per return area based on visitor movement history information for past event groups, and acquires an event group cluster by clustering multiple events that have similar characteristics regarding the number of visitors per return area, and includes a visitor identification unit 11A, a statistical processing unit 11B, and a prediction model learning unit 11C.

[0016] Among these, the visitor identification unit 11A acquires event information on a group of past events from the event DB 14, and identifies visitors to the event related to the acquired event information based on the location information stored in the location information DB 13. For example, a user whose location (mesh code in FIG. 2) during the time period from the opening time to the closing time of a certain event day is a mesh that includes the venue of the event is identified as a visitor to the event. As an example of information on the identified event visitor, as shown in FIG. 4, information on the outbound journey (entrance station, lines 1 and 2, departure time, arrival station, arrival time) and information on the return journey (entrance station, lines 1 and 2, departure time, arrival station, arrival time) are acquired using a user ID that identifies each event visitor as a key.

[0017] The statistical processing unit 11B determines the return area of ​​each visitor after the event from each visitor's movement history on the day of the event, which is obtained based on the location information of the visitor identified by the visitor identification unit 11A, and obtains return area number of people statistical information obtained by statistically analyzing the number of people in each return area and visitor movement history summary information before the start of the event on the day of the event, as "statistical visitor information." As described above, the "return area" is defined as the first staying area in the movement of a visitor immediately after the event, so the statistical processing unit 11B determines the first staying area in the movement immediately after the event (for example, an area where the visitor stayed for a predetermined reference time or more) from each visitor's movement history on the day of the event, and determines the determined staying area as the return area.

[0018] Among the statistical visitor information acquired above, the "return area number of people statistical information" may be, for example, the return area number of people statistical information by city, ward, town, and village as shown in FIG. 5(a), the return area number of people statistical information by station as shown in FIG. 5(b), and the return area number of people statistical information on visitors from far away as shown in FIG. 5(c). Among these, for "visitors from far away," the "number of people" shown in FIG. 5(c) is obtained by identifying visitors whose return area is a candidate facility (e.g., Haneda Airport, Narita Airport, Tokyo Station, etc. as shown in FIG. 5(c)) where visitors from far away are expected to stay on the way home, and counting the number of people for each candidate facility. On the other hand, among the statistical visitor information acquired, the "visitor movement history summary information before the start of the event on the day of the event" may be summary information including the entrance station, arrival station, entrance time, arrival time, number of people, and expanded estimated number of people as shown in FIG. 5(d). In the example information shown in Figures 5(a) to 5(d), the "expanded estimated number of people" means the "number of people in the entire telecommunications industry" that is estimated based on the "number of people" that can be grasped by a certain telecommunications company and the share ratio of that telecommunications company in the telecommunications industry. The "expanded estimated number of people" in the information shown in Figures 6(b) and 7(a) to 7(c) has the same meaning.

[0019] The prediction model learning unit 11C receives as input the visitor movement history summary information from the statistical visitor information acquired by the statistical processing unit 11B and the event information acquired by the visitor identification unit 11A, generates a prediction model that outputs the return area number of people statistical information from the statistical visitor information, and acquires an event group cluster by clustering multiple events having similar characteristics regarding the number of people in each return area for each event in the prediction model using a predetermined clustering method (Ward method, etc.). The generated prediction model is transferred to the prediction unit 12 and stored by the prediction unit 12. The prediction model learning unit 11C updates and brushes up the prediction model 12A stored by the prediction unit 12 with each subsequent learning.

[0020] Next, the prediction unit 12 is a functional unit that acquires a regression result of the prediction of the number of visitors of the target event for each return area from the event information of the target event input from the business terminal 30 using a regularized regression method (e.g., ridge regression, lasso regression, etc.) while referring to the prediction model 12A, identifies a nearest cluster that is closest to the regression result among the event group clusters acquired by clustering by the cluster acquisition unit 11 based on the acquired regression result, predicts the number of visitors of the target event for each return area based on at least the center of gravity of the nearest cluster using various methods described later, and outputs the predicted value to the business terminal 30. Regarding the identification of the nearest cluster, for example, the prediction unit 12 identifies the event group cluster whose center of gravity of each event group cluster is closest to the position indicated by the regression result of the prediction of the number of visitors of the target event for each return area as the nearest cluster.

[0021] Next, the process executed in the return area prediction device 10 will be described with reference to the flow diagrams of Fig. 8(a) to Fig. 8(c). In the location information collection process shown in Fig. 8(a), in the return area prediction device 10, location information transmitted from the user terminal 20 is received and stored in the location information DB 13 as needed (step S0). Similarly, although not shown in the figure, past event information transmitted from the business terminal 30 is received and stored in the event DB 14 as needed.

[0022] The prediction model learning process shown in FIG. 8(b) is started, for example, when the past event information is updated as a trigger. First, the visitor identification unit 11A acquires event information related to a group of past events from the event DB (step S1), and identifies visitors to the event related to the acquired event information based on the location information stored in the location information DB (step S2). The location information of the identified visitors is transferred to the statistical processing unit 11B, which acquires the movement history of each visitor on the day of the event from the location information of the identified visitors (step S3), determines the return area of ​​each visitor after the event from the movement history of each visitor on the day of the event, and acquires statistical visitor information by statisticalizing the number of people for each return area (step S4). The acquired statistical visitor information is transferred to the prediction model learning unit 11C.

[0023] The prediction model learning unit 11C receives as input the visitor movement history summary information from the acquired statistical visitor information and the event information acquired by the visitor identification unit 11A, generates a prediction model that outputs the return area number of people statistical information from the statistical visitor information, and stores it in the prediction unit 12 as the prediction model 12A (step S5). Alternatively, the prediction model learning unit 11C updates and brushes up the prediction model 12A stored by the prediction unit 12 after each subsequent learning.

[0024] Furthermore, the prediction model learning unit 11C obtains an event group cluster by clustering a plurality of events having similar characteristics regarding the number of people in each return area for each event of the prediction model using a predetermined clustering method (Ward method, etc.) (step S6). For example, as shown in FIG. 9(a), a cluster X consisting of events 1 and 2, a cluster Y consisting of event 3, and a cluster Z consisting of events 4 and 5 are obtained as an event group cluster. Note that FIG. 9(a) is a graph in which the number of people information for each return area in various events is displayed on a two-dimensional coordinate system consisting of an axis of the number of people in the return area (Adachi Ward) and an axis of the number of people in the return area (Shibuya Ward) for convenience in order to make it easier to understand, and in reality, there is an axis for the number of cities, wards, towns, and villages to be analyzed. The same applies to the graphs of FIG. 9(b), FIG. 9(c), FIG. 10(b), and FIG. 11(b) described later.

[0025] The return area prediction process shown in Fig. 8(c) is started, for example, when an operator inputs an execution start command together with the event information of the target event from the business terminal 30. Examples of the event information of the target event include the information exemplified in Fig. 6(a) and Fig. 6(b), where the data format of Fig. 6(a) is the same as the example of past event information in Fig. 3 described above, and the data format of Fig. 6(b) is the same as the example of information in Fig. 5(d) described above.

[0026] First, the prediction unit 12 acquires event information of a target event from the business terminal 30 (step S11), and acquires a regression result of the predicted number of visitors to the target event for each return area from the event information of the target event using ridge regression or the like (step S12). Then, the prediction unit 12 compares the distance between the center of gravity of each event group cluster and the position indicated by the regression result (step S13). In this comparison, the event group cluster with the shortest distance is identified as the nearest cluster. For example, as shown in FIG. 9(b), the distance between the center of gravity of each event group cluster X, Y, and Z and the position indicated by the regression result of event A is compared, and the event group cluster X with the shortest distance is identified as the nearest cluster.

[0027] Furthermore, as an example of a prediction method, the prediction unit 12 predicts the number of people represented by the center of gravity of the nearest cluster as the number of people for each return area of ​​the target event (step S14), and outputs the prediction result to the business terminal 30 (step S15). For example, as shown in FIG. 9(c), the number of people represented by the center of gravity of the event group cluster X, which is the nearest cluster, is set as the predicted number of people for each return area of ​​the target event (event A). Examples of the predicted number of people for each return area of ​​the target event (event A) include the return area prediction result for each city, ward, town, and village shown in FIG. 7(a), the return area prediction result for each station shown in FIG. 7(b), and the return area prediction result for visitors from far away shown in FIG. 7(c). Note that the "output" in step S15 includes various forms of output, such as display output on a display, print output on a printer, and data transmission to an external device.

[0028] The location information of the user terminal 20 used in the above embodiment is not limited to connection information of the user terminal 20 to a base station, but may also be GPS location information of the user terminal 20, Wi-Fi (registered trademark) connection information of the user terminal 20, Bluetooth (registered trademark) beacon connection information, etc., and similar results can be obtained.

[0029] The effects of the above embodiment will be described below.

[0030] The prediction unit 12 predicts the number of people in each return area of ​​the target event based on at least the centroid position of the identified nearest cluster. In this way, the nearest cluster that is closest to the regression result by the regularized regression method is identified, and the number of people in each return area of ​​the target event is predicted based on at least the centroid position of the nearest cluster. This makes it possible to accurately predict the number of people in each return area of ​​the target event even if the event has never been held in the past or has only a small number of events held in the past, or even if tickets have not yet been released.

[0031] As an example of a prediction method, the prediction unit 12 predicts the number of people represented by the center of gravity of the nearest cluster as the number of people for each return area of ​​the target event. When predicting the number of people for each return area of ​​the target event in this way, the center of gravity of the nearest cluster is used as basic information. However, as the basic information at this time, in addition to the center of gravity of the nearest cluster, a regression result by a regularized regression method may be used as well, as in Modifications 1 and 2 described below. However, it is avoided to use only the regression result. This prevents overlearning, which is too faithful to the regression result obtained by regression from the event information of the target event and deviates from the original data trend, and avoids a situation in which the prediction accuracy of the return area is hindered.

[0032] 8(b), the cluster acquisition unit 11 generates a prediction model using the visitor movement history summary information from the statistical visitor information and the event information acquired by the visitor identification unit 11A as input, and outputting the return area number of people statistical information from the statistical visitor information, and acquires an event group cluster by clustering a plurality of events having similar characteristics regarding the number of people in each return area for each event in the prediction model using a clustering method such as Ward's method. By generating a prediction model in this way and using a method of clustering a plurality of events having similar characteristics regarding the number of people in each return area for each event in the prediction model using a clustering method such as Ward's method, it is possible to acquire an event group cluster that fully reflects the characteristics regarding the number of people in each return area for past events, which can contribute to improving the prediction accuracy of the return area.

[0033] (Variation 1 regarding prediction of the number of people per return area of ​​the target event) In the above embodiment, an easily realizable processing example has been described in which the prediction unit 12 predicts the number of people represented by the center of gravity of the nearest cluster as the number of people for each return area of ​​the target event. However, this is only one example of a prediction method, and prediction methods other than the above (Modifications 1 and 2) as described below may also be adopted.

[0034] As a first modification, the prediction unit 12 may predict the number of people represented by the midpoint between the centroid position of the nearest cluster and the position indicated by the regression result as the number of people for each return area of ​​the target event. Specifically, in step S14A of the return area prediction process shown in FIG. 10(a), the prediction unit 12 obtains the midpoint between the centroid position of the nearest cluster identified from the comparison result of step S13 and the position indicated by the regression result, and predicts the number of people represented by the midpoint as the number of people for each return area of ​​the target event. In terms of the two-dimensional coordinate system shown in FIG. 10(b), the distance between the position indicated by the regression result of the target event A and the centroid positions of the event group clusters X to Z is compared, and since the distance to the event group cluster X is the shortest, the event group cluster X is identified as the nearest cluster, and the number of people represented by the midpoint P between the centroid position of this nearest cluster and the position indicated by the regression result is predicted as the number of people for each return area of ​​the target event. In this type of variant example 1, the regression results are used as part of the basic information for prediction, but the number of people represented by the midpoint P between the center of gravity of the nearest cluster and the position indicated by the regression results is predicted as the number of people in each destination area of ​​the target event. Therefore, the regression results are appropriately taken into account, overlearning is prevented, and the number of people in each destination area of ​​the target event can be appropriately predicted.

[0035] (Variation 2 regarding prediction of the number of people per return area of ​​the target event) Next, the second modification will be described with reference to Figures 11(a) and (b). In step S14B of the return area prediction process shown in Figure 11(a), the prediction unit 12 judges whether or not the position indicated by the regression result exists within the boundary of the nearest cluster identified from the comparison result of step S13. Here, as shown in "Event B Regression Result" in Figure 11(b), if the position Q indicated by the regression result of the target event B exists within the boundary of the nearest cluster (YES in step S14B), the prediction unit 12 predicts the regression result as the number of people for each return area of ​​the target event B (step S14C).

[0036] On the other hand, in step S14B, if the position indicated by the regression result of the target event C is not within the boundary of the nearest cluster, as shown in "Event C Regression Result" in Figure 11(b) (NO in step S14B), the prediction unit 12 predicts the number of people represented by the intersection R of the line connecting the position indicated by the regression result and the center of gravity of the nearest cluster (cluster Z) with the boundary line of the nearest cluster (cluster Z) as the number of people in each return area of ​​the target event (step S14D).

[0037] In this type of modification 2, the regression result is used as part of the basic information for prediction, but when the position indicated by the regression result of the target event is within the boundary of the nearest cluster, that is, when the regression result is an appropriate value (not inappropriate enough to be considered as overlearning), the appropriate regression result can be predicted as the number of people for each return area of ​​the target event. On the other hand, when the position indicated by the regression result of the target event is not within the boundary of the nearest cluster, the number of people indicated by the intersection R is predicted as the number of people for each return area of ​​the target event, thereby preventing the prediction result from deviating into the outer area of ​​the nearest cluster (cluster Z) and preventing the value from becoming inappropriate enough to be considered as overlearning.

[0038] (Explanation of terms, explanation of hardware configuration (Fig. 12), etc.) The block diagrams used to explain the above embodiments and modifications show functional blocks. These functional blocks (components) are realized by any combination of at least one of hardware and software. The method of realizing each functional block is not particularly limited. That is, each functional block may be realized by using one device that is physically or logically combined, or may be realized by using two or more devices that are physically or logically separated and directly or indirectly connected (for example, by wire, wirelessly, etc.). The functional blocks may be realized by combining the one device or the multiple devices with software.

[0039] Functions include, but are not limited to, judgment, determination, judgment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs the function of transmission is called a transmitting unit or a transmitter. As described above, there is no particular limitation on the method of realization of each of these.

[0040] For example, the return area prediction device in an embodiment of the present disclosure may function as a computer that performs the processing in this embodiment. Fig. 12 is a diagram showing an example of a hardware configuration of a return area prediction device 10 according to an embodiment of the present disclosure. The above-mentioned return area prediction device 10 may be physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, and the like.

[0041] In the following description, the term "apparatus" may be replaced with a circuit, a device, a unit, etc. The hardware configuration of the destination area prediction device 10 may be configured to include one or more of the devices shown in the drawings, or may be configured to exclude some of the devices.

[0042] Each function in the return area prediction device 10 is realized by loading a specific software (program) onto hardware such as a processor 1001 and a memory 1002, causing the processor 1001 to perform calculations, control communication by a communication device 1004, and control at least one of reading and writing of data in the memory 1002 and the storage 1003.

[0043] The processor 1001 controls the entire computer by running, for example, an operating system. The processor 1001 may be configured with a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, and the like.

[0044] Furthermore, the processor 1001 reads out programs (program codes), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 into the memory 1002, and executes various processes according to these. As the programs, those that cause a computer to execute at least a part of the operations described in the above-mentioned embodiments are used. Although the above-mentioned various processes have been described as being executed by one processor 1001, they may be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented by one or more chips. The programs may be transmitted from a network via a telecommunications line.

[0045] The memory 1002 is a computer-readable recording medium, and may be configured by at least one of, for example, a Read Only Memory (ROM), an Erasable Programmable ROM (EPROM), an Electrically Erasable Programmable ROM (EEPROM), a Random Access Memory (RAM), etc. The memory 1002 may be called a register, a cache, a main memory (primary storage device), etc. The memory 1002 can store a program (program code), software modules, etc. that are executable to implement a wireless communication method according to an embodiment of the present disclosure.

[0046] Storage 1003 is a computer-readable recording medium, and may be composed of at least one of, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray (registered trademark) disk), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy (registered trademark) disk, a magnetic strip, and the like. Storage 1003 may be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database including at least one of memory 1002 and storage 1003, or other suitable medium.

[0047] The communication device 1004 is hardware (transmission / reception device) for performing communication between computers via at least one of a wired network and a wireless network, and is also called, for example, a network device, a network controller, a network card, or a communication module.

[0048] The input device 1005 is an input device (e.g., a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that accepts input from the outside. The output device 1006 is an output device (e.g., a display, a speaker, an LED lamp, etc.) that performs output to the outside. The input device 1005 and the output device 1006 may be integrated (e.g., a touch panel). Each device such as the processor 1001 and the memory 1002 is connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or may be configured using different buses between each device.

[0049] Each aspect / embodiment described in this disclosure may be used alone, in combination, or switched according to execution. In addition, notification of specific information (e.g., notification that "X is the case") is not limited to being done explicitly, but may be done implicitly (e.g., not notifying the specific information).

[0050] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described in the present disclosure. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0051] The order of the steps, sequences, flow charts, etc. of each aspect / embodiment described in this disclosure may be changed unless inconsistent. For example, the methods described in this disclosure present elements of various steps using an example order, and are not limited to the particular order presented.

[0052] The input and output information may be stored in a specific location (e.g., memory) or may be managed using a management table. The input and output information may be overwritten, updated, or added to. The output information may be deleted. The input information may be transmitted to another device.

[0053] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0054] When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, similar to the term "comprising." Further, when used in this disclosure, the term "or" is not intended to be an exclusive or.

[0055] In this disclosure, where articles have been added due to translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0056] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may also be interpreted in the same way as "different." [Explanation of symbols]

[0057] 10...return area prediction device, 11...cluster acquisition unit, 11A...visitor identification unit, 11B...statistical processing unit, 11C...prediction model learning unit, 12...prediction unit, 12A...prediction model, 13...location information DB, 14...event DB, 20...user terminal, 30...business terminal, 1001...processor, 1002...memory, 1003...storage, 1004...communication device, 1005...input device, 1006...output device, 1007...bus.

Claims

1. a prediction unit that uses a regularized regression method to obtain a regression result of predicting the number of visitors of a target event for each return area from event information of the target event, identifies a nearest cluster that is closest to the regression result among the event group clusters that have already been clustered based on the obtained regression result, and predicts the number of visitors of the target event for each return area based on at least the center of gravity of the nearest cluster; A return area prediction device comprising:

2. a cluster acquisition unit that acquires the event group cluster by determining the number of visitors to each event for each return area based on movement history information of visitors to past events, and clustering a plurality of events having similar characteristics regarding the number of visitors to each return area; The return area prediction device according to claim 1 , further comprising:

3. The cluster acquisition unit a visitor identification unit that acquires event information related to a group of past events and identifies visitors to an event related to the acquired event information based on location information stored in a location information database that stores location information of various users; a statistical processing unit that obtains a return area for each visitor after the event from the movement history of each visitor on the day of the event, which is obtained based on the location information of the visitor identified by the visitor identification unit, and obtains, as statistical visitor information, return area number of people statistical information obtained by statistically calculating the number of people for each return area and visitor movement history summary information before the start of the event on the day of the event; a prediction model learning unit that receives as input the visitor movement history summary information and the event information acquired by the visitor identification unit from the statisticalized visitor information acquired by the statistical processing unit, generates a prediction model that outputs the return area number of people statistical information from the statisticalized visitor information, and acquires the event group clusters by clustering a plurality of events having similar characteristics regarding the number of people per return area for each event in the prediction model using a predetermined clustering method; The return area prediction device according to claim 2 , further comprising:

4. The prediction unit is The number of people represented by the center of gravity of the nearest cluster is predicted as the number of people for each destination area of ​​the target event. The return area prediction device according to claim 1 .

5. The prediction unit is predicting the number of people represented by the midpoint between the center of gravity of the nearest cluster and the position indicated by the regression result as the number of people for each destination area of ​​the target event; The return area prediction device according to claim 1 .

6. The prediction unit is If the location indicated by the regression result is within the boundary of the nearest cluster, the regression result is predicted as the number of people for each destination area of ​​the target event; If the position indicated by the regression result is not within the boundary of the nearest cluster, the number of people indicated by the intersection of a straight line connecting the position indicated by the regression result and the center of gravity of the nearest cluster with the boundary line of the nearest cluster is predicted as the number of people for each destination area of ​​the target event. The return area prediction device according to claim 1 .

Images