JP5707033B2 - Behavior trend analysis system and behavior trend analysis method - Google Patents

Description

  The present invention relates to a method and an apparatus for analyzing a behavior tendency of a user of a station or a nearby commercial facility from a use history of card or terminal operation.

  Advances in information and communication technology will record electronic behavior records (use history), and by analyzing the use history, predict what kind of behavior the person will have It became possible to do.

  For example, at a convenience store, products sold from cash register terminals and their purchasers (gender / rough age) and time are collected, data is analyzed using a POS system (Point of sale system), and products that are in the customer base at each store. It is well known that sales per store area can be improved by displaying at the appropriate time.

  In the above example, the analysis for identifying the individual buyer is not performed, but the Web commerce site on the Internet can grasp not only the purchase of the product but also the information of the referenced product as the Web operation history. Therefore, it becomes possible to analyze the preference for each deeper user. For this reason, a technique called so-called recommendation is adopted, in which products that are likely to meet the user's interest are presented as “recommendations” from past usage records for each user. This recommendation technique can also be applied to IC cards, mobile phones and the like that can collect information electronically online.

  In this way, with the widespread use of devices that can electronically record and collect purchasing information, the optimization of product assortment by analysis of conventional purchase behavior records that do not identify individuals, and the purchase behavior history of each individual More advanced marketing techniques that presume consumer preferences, present recommended products optimized for each individual, and promote purchases are becoming widespread.

  Patent Document 1 describes an invention for improving the quality of recommendation. Patent Document 2 describes an invention related to recommendation based on user behavior characteristics.

JP 2006-318294 A JP 2004-102950 A

  In the recommendation, if the amount of information given to the user is too much, or if the information is unnecessary information (poor quality) for the user at that time, presenting the information itself is uncomfortable to the user It becomes a factor to give. Therefore, in order to improve the quality of the recommendation, a method for improving the quality by using the place and frequency such as repeatedly purchasing the same item at the same place has been proposed. However, recommendations that depend on individual preferences are not necessarily beneficial to the user. For example, even if information on other weekly magazines is provided to a user who buys a specific weekly magazine regularly at a station, it is not necessarily useful information for the user.

  Furthermore, there is a restriction in the case of a contactless IC card or a mobile phone having an equivalent function that is rapidly spreading as an electronic ticket or a small payment method. That is, it is anonymity of a non-contact type IC card or a mobile phone having an equivalent function. In fact, with a registered card, an individual can be identified from its ID number, but from the user's point of view, the card or mobile phone can be used without handing over a card with a name written like a credit card or signing the name. The fact that payment can be completed simply by holding it over a reader / writer device for payment is behind the rapid spread. Therefore, a recommendation based on an excessive personal history may cause discomfort to the user.

  In view of such circumstances, the present invention is not limited to information that is excessively dependent on the individual purchase history of the user, such as recommendations for purchasing behavior in or around the station without giving the user discomfort. The purpose is to provide information.

  In the present invention, like the conventional recommendation, the personal preference is analyzed from the viewpoint of how to use the station, not from the viewpoint of what and where to purchase frequently. In a non-contact type IC card or a mobile phone having an equivalent function, it is considered that a recommendation considering anonymity is necessary. Therefore, providing information without making the user uncomfortable with information that is excessively dependent on the individual purchase history of the user by indexing individual users as station user images and making recommendations Can do.

  More specifically, by using the non-contact traffic IC card possessed by each user as a transportation settlement means or the use history of a mobile phone having an equivalent function, Classify usage by three axes. The three axes are, first, whether the user's tendency to use the station is for convenience or entertainment, second, the frequency of use of the user, and third, the user's station. Staying time at. By classifying on the basis of these axes, it is possible to provide information that is most suitable for the station usage mode directed by the user.

  The station is a base for traveling by rail, and has various uses such as purchasing, dining, time adjustment and waiting in and around the station. By using the classification of station usage forms according to the present invention, it becomes possible to provide information on similar stores, products, or services based on how the user is using daily. Therefore, it is possible to provide information without any discomfort based on the individual preference of the purchased item.

It is an example of the block diagram of the whole system of this embodiment. It is a figure explaining an example of the procedure of the total process in a calculation server. It is a figure explaining an example of the structure of the record which stores one use history of a non-contact IC card. It is a figure explaining an example of the structure of the index value which shows a user's action tendency. It is a figure explaining an example of the structure of the simple index value which shows the user's action tendency recorded in a non-contact IC card. It is a figure explaining an example of the structure of the basic information recorded in a non-contact IC card. It is a figure explaining an example of the procedure of the index value calculation in a calculation server. It is a figure explaining an example of the estimation method of a utilization station, and the estimation method of residence time. It is an example of the figure explaining the basic principle of the stay time estimation method of a station. It is a figure explaining an example of the structure of the personal setting file hold | maintained for every user in calculating | requiring an index value. It is a figure explaining an example of the structure of the store data stored in a partner store advertisement file. It is a figure explaining an example of the method of estimating the correction value of the residence time in a station premises. It is a figure explaining an example of the method of estimating the correction value of the residence time around a station. It is a figure explaining an example of the structure of the advertisement data stored in a partner store advertisement file. It is a figure explaining an example of composition of an analysis information screen for partners. It is a figure explaining an example of composition of a sales promotion information screen for partners. It is a figure explaining an example of composition of an offer information screen for users. It is a figure explaining an example of the structure of the provision information screen for users for mobile phones. It is a figure explaining an example of composition of an advertising information screen for users.

  A method for classifying station user usage forms according to the present invention with reference to FIGS. 1 to 19, and a company operating a station user and a commercial facility in or around the station using the classification (hereinafter referred to as a partner). An example of a system for providing information to is shown.

  FIG. 1 is a block diagram of a system for implementing the present invention. The main components of the system are contactless IC cards (or mobile phones 1021, 1022, 1023, and 1024 incorporating an equivalent function) directly used by users (1011, 1012, 1013, 1014, and 1015 in FIG. 1). 1025), terminal devices (ticket gates 1031, ticket vending machines 1032 and 1033) for railway use, which are arranged on the premises, around the station, or in stores outside the city, and for payments related to purchase of goods and service benefits Non-contact IC card incorporated in a terminal device (POS cash register terminal 1035, information terminal, etc. 1034) (hereinafter referred to as a non-contact IC card including a portable information device such as a mobile phone incorporating a non-contact IC card function) Reader / writer devices (1041, 1042, 1043, 1044, 1045), reader / writer devices and data storage Analysis processing is a network 1071 for connecting the server group (1081,1082,1083,1084) or the like is performed.

  Here, the non-contact IC card 1021 includes a CPU 1111 that controls the function of the card and processes data, a nonvolatile memory 1121 that records data, a wireless module 1131 that supplies power using communication and radio wave energy, and a wireless module. It consists of an antenna 1141. This structure is the same in 1022, 1023, 1024, 1025, CPU 1112, 1113, 1114, 1115, nonvolatile memory 1122, 1123, 1124, 1125, wireless modules 1132, 1133, 1134, 1135, antennas 1142, 1143, 1144, 1145. Note that functions (for example, a power supply circuit and an encryption circuit) in the non-contact IC card that are not directly related to the description of the present invention are omitted in FIG.

  Similarly, the reader / writer device (hereinafter referred to as reader / writer device) 1041 of the non-contact IC card controls the function of the device, processes the data, and controls the communication performed via the network 1071, holds the data to be written to the IC card. And a wireless module 1231 for communicating with the IC card, and a wireless antenna 1241. This structure is the same also in 1042, 1043, 1044, 1045, CPU1212,1213,1214,1215, memory 1222,1223,1224,1225, wireless module 1232,1233,1234,1235, antenna 1242,1243,1244,1245. Consists of. Furthermore, the reader / writer device is equipped with one or two display devices (1051, 1052, 1053, 1054, and 1063, 1065), and information can be presented to the user and the store clerk. Each reader / writer device (1041, 1042, 1043, 1044, 1045) is connected to the network 1071, and the information of the read contactless IC card is added to the information of the store or device where the reader / writer device is installed, etc. Information to be transmitted to the history server 1081 and written to the non-contact IC card can be calculated by the internal CPU (1211, 1212, 1213, 1214, 1215) or received via the network 1071. Furthermore, information generated based on values calculated by the CPU (1211, 1212, 1213, 1214, 1215) and information received via the network 1071 are displayed on the display devices (1051, 1052, 1053, 1054, and 1063, 1065). Can be displayed. It should be noted that the functions of a terminal device (for example, a gate control mechanism of a ticket gate or a product barcode reading mechanism of a POS cash register terminal) having a built-in card reader device that is not directly related to the description of this embodiment are omitted in FIG. is there.

  The server group includes a history server 1081, a calculation server 1082, a partner information distribution server 1083, and a user information server 1084. These servers are connected to the network 1071 and can communicate with each other or with each reader / writer device (1041, 1042, 1043, 1044, 1045).

  The history server 1081 includes a network interface, a processor, a memory, and a recording device (not shown). The network interface is an interface for connecting to a network. The recording device records each program and each data, and is, for example, a hard disk drive, a CD-ROM drive, or a flash memory. Each program and each data may be divided and recorded on a plurality of recording devices. The processor implements various functions by reading each program recorded in the recording device into a memory and executing the program. Specifically, by executing a data collection program, the use history data of station users and store users read by the reader / writer device (1041, 1042, 1043, 1044, 1045) is collected through the network 1071, and the server To the internal recording device. The format of one record of usage history data to be collected is shown in FIG. A record consists of nine fields. Who represents the record (ID field 301 for storing the user ID), when (date / time field 302 in the year / month / day / hour / minute format for storing the use date / time), where (the station code identifying the station, or the store Station store field 303 for storing the store code for identifying the terminal and terminal field 304 for storing the identification code of the station or terminal in the store), what was done (type field 305 for storing the code indicating the type of use) , How much (amount field 306 for storing usage amount and information used as a calculation reference thereof, balance 307, calculation reference 1 field 308, calculation reference 2 field 309). Here, in the calculation standard 1 and calculation standard 2 fields, for example, the station code of the boarding station and the getting-off station, which are the calculation standards of the railway fare, the product code indicating the type of the purchased product and the received service, and the like are stored.

  The calculation server 1082 calculates the classification that is a feature of the present embodiment for each station user. Details will be described later. The calculation server includes a network interface (I / F (A) and I / F (B)) 1311, a processor 1312, a memory 1313, and a recording device 1314. The network interface is an interface for connecting to a network. The recording device records various programs and various data, such as a hard disk drive, a CD-ROM drive, and a flash memory. Various programs and various data may be divided and recorded on a plurality of recording devices.

  The processor implements various functions by reading various programs recorded in the recording device into a memory and executing them. The result of the processing for calculating the classification index for each user is stored in the user classification index file in the recording device. In addition, information on a store operated by the partner and advertisement information on the store are stored in a partner store advertisement file in the recording device. This partner store advertisement file can be referred to, added, and changed by the partner person 1331 from the partner company terminal 1321 through the Internet 1072.

  The partner information distribution server 1083 includes a network interface, a processor, a memory, and a recording device (not shown). The network interface is an interface for connecting to a network. The recording device records each program and each data, and is, for example, a hard disk drive, a CD-ROM drive, or a flash memory. Each program and each data may be divided and recorded on a plurality of recording devices. The processor implements various functions by reading each program recorded in the recording device into a memory and executing the program. Specifically, the data stored in the user classification index file is totaled and accumulated. This information can be referred from the partner company terminal 1321 through the Internet 1072. Details of the information generated here will be described later.

  The user information server 1084 includes a network interface, a processor, a memory, and a recording device (not shown). The network interface is an interface for connecting to a network. The recording device records each program and each data, and is, for example, a hard disk drive, a CD-ROM drive, or a flash memory. Each program and each data may be divided and recorded on a plurality of recording devices. The processor implements various functions by reading each program recorded in the recording device into a memory and executing the program. Specifically, the data stored in the user classification index file is aggregated, information for individual users is generated, and distributed to individual users as Web or mobile phone mail (1411, 1412). Web information is delivered to each user 1441 through the Internet 1072 through a personal computer 1431 installed at each user's home. On the other hand, the mail information is delivered to each user 1442 as a mail through the line of the mobile phone carrier 1073 to the information terminal 1432 such as a mobile phone owned by each user.

  In this embodiment, the server group of the history server 1081, the calculation server 1082, the partner information distribution server 1083, and the user information server 1084 is described. However, the functions of these server groups can be executed by one or a plurality of servers. It goes without saying that it is also possible to do this.

  From the viewpoint of the user using this system, there are two major points of contact. One is to touch a reader / writer device with a non-contact IC card when using a railway in a station or a commercial facility around the station, making a payment at a store, or using an information terminal. All touched records are sent to the history server 1081 and recorded. Further, information on the station usage status of the person himself / herself, information on the railway, station, and commercial facilities around the station is obtained from a display device connected to the reader / writer device. The other is to obtain information on the station usage status of the person himself / herself, the railway, the station, and the commercial facilities around the station from the personal computer 1431 and the information terminal 1432 owned by the user.

  On the other hand, there are two major points of contact when viewed from the perspective of partner companies using this system. One is to use information about individual users displayed on the display devices (1063, 1065) connected to the reader / writer device when customer service is performed at a store operated by a partner company in a commercial facility around the station. That is. The other is to confirm the trend of the user with respect to the advertisement information presented by the partner from the terminal 1321.

  Next, details of functions of the reader / writer device (1041, 1042, 1043, 1044, 1045), calculation server 1082, partner information distribution server 1083, and user information server 1084 will be described in order.

  When the non-contact IC card is touched, the reader / writer device performs a predetermined operation by the function of the terminal. In the non-contact IC card non-volatile memory 1121, 1122, 1123, 1124, 1125, as shown in FIG. 6, the card-specific ID 601 and the balance 602, the station code 603 information of the station that finally passed through the ticket gate, The most recent constant of 3 history records (for example, 20 records depending on the memory capacity) and an index value to be described later are recorded.

  In the case of boarding, the ticket gate 1031 first reads the ID 601 and balance 602 of the card and writes the station code of the boarding station into the station code 603 of the card. ID 301 from which ID 301 is read, date and time 302 are the date and time at that time, station store 303 is the station code, terminal 304 is the ticket gate code, type 305 is “ticket gate”, amount 306 is “0”, and balance 307 is the balance 602 that has been read. The calculation criteria 1 (308) and the calculation criteria 2 (309) are both blank, and the record shown in FIG. At the same time, a record having the same content is sent to the history server 1081. Further, distribution information for the user corresponding to the ID 601 of the card is acquired from the user web 1411 of the user information server 1084 and displayed on the display device 1051. Details of the information to be displayed will be described later. Further, an index value (in this case, a simple index for a non-contact IC card) indicating the user's behavior tendency corresponding to the ID 601 of the card is read from the user classification index file of the calculation server 1082, and the index of this non-contact IC card is read. Write to the value storage part (update the value).

  In case of getting off, the ID 601 of the card, the balance 602, and the immediately preceding boarding station code 603 are read, and the station code 603 is deleted (a code indicating a blank is written). First, a predetermined fare is calculated from the boarding station code 603 and the station code where the ticket gate is installed, and the amount F is obtained. Next, IDs 601 and 302 from which the ID 301 is read out are the date and time at that time, the station store 303 is the station code, the terminal 304 is the ticket gate code, the type 305 is the “ticket gate”, the amount 306 is first obtained, and the balance 307 is The value obtained by subtracting F from the balance 602 that has been read, calculation standard 1 (308) and calculation standard 2 (309) are the two station codes (the boarding station code 603 and this ticket gate) that were previously used for calculating the fare. 3 is prepared and written in the history record recording area of the non-contact IC card. At the same time, the balance 602 is rewritten to the same value as the balance 307. At the same time, a record having the same content is sent to the history server 1081. Further, distribution information for the user corresponding to the ID 601 of the card is acquired from the user web 1411 of the user information server 1084 and displayed on the display device 1051. Further, an index value (in this case, a simple index for a non-contact IC card) representing the user's behavior tendency corresponding to the ID 601 of the card is read from the user classification index file of the calculation server 1082, and the index of this non-contact IC card is read. Write to the value storage part (update the value).

  The ticket vending machine 1032, the POS cash register terminal 1035, or the terminal 1033 for purchase settlement such as a ticket vending machine or a vending machine first reads the ID 601 and the balance 602 of the card. Next, IDs 601 and 302 from which the ID 301 is read out are the date and time at that time, the station store 303 is the station code, the terminal 304 is the terminal code, the type 305 is “passenger ticket”, “product sales”, “service”, etc. The amount 306 is the sales amount, the balance 307 is a value obtained by subtracting the sales amount from the read balance 602, and the calculation criteria 1 (308) and calculation criteria 2 (309) are items of sale items, product codes, etc. 3 is produced and written in the history record recording area of the non-contact IC card. At the same time, the balance 602 is rewritten to the same value as the balance 307. At the same time, a record having the same content is sent to the history server 1081. Further, distribution information for the user corresponding to the ID 601 of the card is acquired from the user web 1411 of the user information server 1084 and displayed on the display device 1052 or 1055. At the same time, distribution information for the user corresponding to the ID 601 of the card is acquired from the partner information distribution server 1083 and displayed on the display devices 1063 and 1065. Details of the information to be displayed will be described later. Further, an index value (in this case, a simple index for a non-contact IC card) representing the user's behavior tendency corresponding to the ID 601 of the card is read from the user classification index file of the calculation server 1082, and the index of this non-contact IC card is read. Write to the value storage part (update the value). In addition, when the so-called charge is added to the balance of the contactless IC card at the ticket vending machine, the type 305 is “deposit”, the amount 306 is a negative value of the amount of deposit, and the balance 602 and balance 307 are set after the deposit. Value.

  The information terminal 1034 first reads out the ID 601 and the balance 602 of the card. Next, ID 601 from which ID 301 was read out, date and time 302 as date and time at that time, station store 303 as station code, terminal 304 as ticket gate code, type 305 as “information”, amount 306 as “0”, and balance 307 as read The same value as the balance 602, calculation criteria 1 (308) and calculation criteria 2 (309) are both blank, and the record shown in FIG. 3 is created and written in the history record recording area of the non-contact IC card. At the same time, a record having the same content is sent to the history server 1081. Further, distribution information for the user corresponding to the ID 601 of the card is acquired from the user web 1411 of the user information server 1084 and displayed on the display device 1054. Further, an index value (in this case, a simple index for a non-contact IC card) representing the user's behavior tendency corresponding to the ID 601 of the card is read from the user classification index file of the calculation server 1082, and the index of this non-contact IC card is read. Write to the value storage part (update the value).

  As described above, the present embodiment is characterized in that the index value representing the user's behavior tendency is written in the index value storage portion of the non-contact IC card. As a result, even if the reader / writer device is not connected to the user information server online, if the reader / writer device stores distribution information for the user, the non-contact IC card is used as the reader / writer device. When communication is performed, it is possible to display distribution information for users based on the ID of the non-contact IC card.

  The calculation server 1082 analyzes the behavior tendency of each user. The processor 1312 calculates an index value indicating the behavior tendency of each user. First, each field of the index value record will be described with reference to FIGS.

  FIG. 4 shows the detailed index value record structure for one user. The start date 401 records the date and time when index value calculation is started. The station use type number 402 records the number of types of stations used by the user after the start date 401. The station use station code 403 records the station code of the station used by the user in the past. The number of elements is equal to the station use type number 402. Reference numerals 411 to 440 are fields that exist for a total of five types of four stations (hereinafter referred to as “Nawabari 4 stations”) frequently used by the user and the other stations as a whole. Here, the frequently used station is a station that is frequently used or a station that the user wants to use. 411 is a station code, 412 is a station flag, and expresses a station attribute by a bit field. Each field represents a residential area, a commercial area, an office / school district, a transfer station, or a sightseeing spot. Depending on the type of station 411, the field is set to 1 if applicable, and the field is set to 0 if not applicable. To do. This information is predetermined for each station. 413 is the number of times the station has been used since the start date 401 (use is a record of getting on and off or purchasing at the station), and the total stay time 414 is the total time the user stayed at the station , The sum of squared stay times 415 is the sum of the squares of the stay times for each time. Details of the method of calculating the stay time will be described later. The reason for obtaining the sum of squares is to obtain the standard deviation as an index. The number of entertainment stores 416 and the number of convenience stores 417 store the total number of times of use depending on whether the store used at the station is an entertainment store or a convenience store. Here, a store with entertainment is a store that handles highly-preferred products, such as a store that handles high-fashion clothing and stores that sell sweets, etc., and a store that has convenience includes convenience stores and side dishes It is a store that handles everyday items such as stores. Whether a certain store is convenient or entertaining can be obtained by referring to store information in the partner store advertisement file. Amusement use amount 418, convenience use amount 419, amusement amount sum of squares 420, and convenience use amount sum of squares 421 are the sum of the amount paid at a store with entertainment and the amount paid at a store with convenience, respectively. The sum of the sum of the squares of the amount of money paid at a store with entertainment and the sum of the squares of the amount of money paid at a store of convenience.

  The index value records 423 to 440 record the number of uses for each time zone and day of the week. The time zone classifies one day into six. In the morning from the first train (3am) to 9:30 am, in the morning from 9:30 am to 11:30 am, in the afternoon from 11:30 am to 1:30 pm, in the afternoon from 1:30 pm to 5pm Until midnight, evening is from 5pm to 8pm, and evening is from 8pm to the last train. The days of the week are classified into three. First, holidays are classified as holidays. Other Monday to Friday are weekdays and Saturday and Sunday are weekends. 423 to 440 record the number of uses after the use start date 401 in the case of 18 combinations of six types of time and three types of days of the week. 423 is weekday morning, 424 is weekday morning, 425 is weekday noon, 426 is weekday afternoon, 427 is weekday evening, 428 is weekday night, 429 is weekend morning, 430 is weekend morning, 431 is Weekend noon, 432 is weekend afternoon, 433 is weekend evening, 434 is weekend evening, 435 is holiday morning, 436 is holiday morning, 437 is holiday day, 438 is holiday afternoon, 439 is holiday In the evening, 440 is a holiday night. In the present embodiment, the above-described classification based on time and date is performed on the assumption that the station usage tendency varies depending on the date and time. However, the present invention is not limited to this.

  In other records representing the entire station, the station code 411 is blank, and the station flag is the sum of all stations that have been used.

  FIG. 5 shows a simple index value record structure for one user. The same information as this record is also recorded on the non-contact IC card held by the user. Since a non-contact IC card has a small amount of built-in memory, it is necessary to reduce and simplify index items (size). The start date 501 is the same as the start date 401 in FIG. 4, the station use type 502 is the same as the station use type 402, station code 1 (511), station flag 1 (512), station code 2 (513), station flag. 2 (514), station code 3 (515), station flag 3 (516), station code 4 (517), and station flag 4 (518) correspond to 4 records corresponding to 4 stations of Nawabari represented by 411 to 431 in FIG. Station code 411 and station flag 412. The number of uses 519, the total stay time 520, the number of entertainment stores 521, the number of convenience stores 522, the amount of entertainment usage 523, and the amount of usage usage 524 of the five records 411 to 431 in FIG. This is the sum of the sum, the sum of stay time 414, the sum of the number of entertainment stores 416, the sum of the number of convenience stores 417, the sum of the entertainment usage amount 418, and the sum of the convenience usage amount 419.

  Next, the processing flow of the processor 1312 will be described with reference to FIG. This process is performed once a day for all users who have used non-contact IC cards.

  First, for all IDs (UIDi) that have been used, the usage history of the day is listed in order of time (201). Next, for each UIDi, the history for one day is analyzed, and each index value is calculated (202). Then, the index value Xik of each UIDi stored in the user classification index file is updated (203). Further, the partner information distribution server 1083 and the user information server 1084 are notified that the index value of each UIDi has been updated.

  Here, the details of the index value calculation method will be described with reference to FIGS. First, data necessary for calculation will be described. FIG. 10 shows user-specific data for each user. ID 101 is a user's UIDi value, and station codes 102 to 105 are codes of four stations of Nawa Bali that are frequently used by the user or set by the user. 111 is data of a typical value 109 and a deviation 110 of travel time calculated from past history data of two stations (station A 106 and station B 107) of the four stations of Nawa Bali. There is a maximum of 16 data of the same format (including boarding / alighting at the same station).

  FIG. 11 shows store data in and around the station and is stored in the partner store advertisement file. The company code 11001 and the company name 11002 are the name and code of the company that manages the store, the store code 11003, the store name 11004 is the name and code of the store, the store code 11005, and the store name 11006 is the branch name and code of the store. is there. The industry code 11007 is a code representing the business type of the store, the station code 11008 is the code of the nearest station of the store, the payment flag 11009 is 0 if the payment at the store is prepayment, 1 if it is postpay, and the inside / outside flag 11010 is that The flag is 0 when the store is in the station yard (within the ticket gate), 1 when the store is around the station (outside the ticket gate), and the trend value 11011 is 1 if the store handles highly entertaining products. A value indicating the tendency of the store to be 2 if it is a store handling high products, the required time 11012 is the average time spent by one customer at the store, and the average amount 11013 is the average value of the payment of one customer at the store The opening time 11014 is the opening time of the store, and the closing time 11015 is the closing time of the store. The required time 11012 is a value set in advance by a field survey or the like, and the average amount 11013 is a value calculated from past results by a POS system or the like. The number of terminals 11020 is the number of POS cash register terminals owned by the store, and the terminal code 11021 is an identification code of the POS cash register terminal. The ticket gate code 11030 and the required time 11031 are a pair of the required time 11031 from the ticket gate 11030 at the station to the store, and there are as many as the number of ticket gates at the station. These required times are also values measured in advance through field surveys.

  The index value is calculated in the order shown in FIG. First, it is estimated which station the user has used based on the history data of a certain user for one day. When using a railway, the records of the use of a contactless IC card include a record of entering the station A through the ticket gate and a record of entering the station B through the ticket gate (FIG. 8 (a)). 801, 802). In this case, the use stations are the station A and the station B, and the records remain in the use record of the non-contact IC card, so that the fact of use can be easily extracted. However, as shown in FIG. 8 (d), when changing to another route at the same railway company, the train gets on and off at the transfer station C, and the station C is used. There is no record. In this case, candidates for a route from station A to station B are calculated in advance from a route map or the like, and transfer station C and transfer time 808 are estimated. Also, if there is a purchase history of vending machines or stores at stations other than station A and station B after passing through the ticket gate at station A and before passing through the ticket gate at station B, the station It is assumed that the user stayed (FIG. 8 (e)).

  Next, store usage tendency values are calculated (702 in FIG. 7). The following processing is performed on all purchase records (FIG. 3) in one day of history data of a certain user.

  First, the station store 303 of the corresponding record is read, and the store data (FIG. 11) having the store code 11005 corresponding to it is searched. The station code 11008 and the trend value 11011 are read out, and if the station code 11008 corresponds to the station code 411 of the Nawabari 4 station of the index value (FIG. 4), 1 is added to the value of the use count 413 and the tendency value 11011 If the value is 1, if the store with high entertainment is used, 1 is added to the number of entertainment stores 416 of the index value, and the value of the amount 306 of the history record is set to the entertainment usage amount 418. The square value is added to the entertainment sum of squares 420. If the value of the trend value 11011 is 2, assuming that a highly convenient store is used, 1 is added to the convenience store number 417 of the index value, and the value of the amount 306 of the history record is added to the convenience usage amount 419. The square value is added to the convenience sum of squares 421. In addition, referring to the value of the date and time 302, the time zone data (423 to 440) corresponding to the time zone (morning, morning, noon, afternoon, evening, night) and the day of the week (weekday, weekend, holiday) determined from the date. Add one.

  If the above station code 11008 does not correspond to the station code of Nawa Bali 4 station, the same processing as above is performed as other stations. Further, the station use station code 403 of the index value is checked, and if the same station code as the station code 11008 does not exist, 1 is added to the station use type 402 and the station code 11008 is added to the station use station code 403.

  Next, the stay time is estimated (703 in FIG. 7). Estimated by dividing into residence time in the station yard (within the ticket gate) and residence time around the station (outside the ticket gate), and the total is taken as the residence time.

  The stay in the station is calculated using the statistical information of the usage history. When entering through the ticket gate of station A and entering through the ticket gate of station B, if there is no stagnation in the station, as shown in FIG. 8 (a), the elapsed time 803 is from the ticket gate to the home. It is the total of walking movement + train waiting time 805, train moving time 804, walking moving time 806 to the ticket gate. When staying in the station yard, a residence time 807 is added to this (FIGS. 8B and 8C). The travel time of the train can be calculated in advance from the time schedule, and the average waiting time can also be calculated from the time schedule, but the walking travel time varies depending on the person. Therefore, the relationship between the required time and the number of users for each time zone of the user who has moved from station A to station B is obtained as a statistical value.

  FIG. 9A plots the required time 901 on the horizontal axis and the number of people per hour (frequency 902) on the vertical axis. There is a limit value (the shortest required time 904) that the train travel time and the train travel time are not physically faster than that (depending on the upper limit of the train operation speed and the walking / running speed of people). In addition, the distribution has a peak centering on people walking at an average speed (average time required 903). When staying at the station (905), it takes longer than the average required time 903. This distribution concept holds true for individuals (FIG. 9B). From an individual perspective, the distribution exists in a range larger than the minimum required time 904, and can be peaked in the same manner as the overall statistical value. When the minimum value 907 and the peak value 906 of the individual distribution are set, a value (908) that becomes 906 + 911 (FIG. 9B) 912 is set as the travel time of the person with respect to the difference 911, and it actually takes The difference (914) obtained by subtracting the value of 906 + 911 from the time 913 is regarded as the residence time in the station. Here, 906 is a typical value, 912 is a time deviation, and this value is obtained in advance corresponding to station A-station B and stored in the user classification index file, so that the residence time estimation criterion for each user can be obtained. To do. If the amount of past data is insufficient and it is difficult to obtain these values, use the values for the entire user (FIG. 9A) (typical value is 903, time deviation) 903-904).

  The residence time obtained above is an estimate of the total residence time in the station, and the distribution among the entrance station, the departure station, and the transfer station is not known. Therefore, it is considered that all the residence times have passed at the station with the highest usage frequency (usage frequency) of the user among the entrance station, the departure station, and the transfer station.

If there is a record of purchase with a non-contact IC card between admission and entry, correction is made based on that time. An example of this correction method will be described with reference to FIGS. In the present embodiment, the data accuracy is improved by correcting the station stay time based on the purchase record included in the usage history.
(A) When there is a usage record at the entrance station: If the latest usage time is used instead of the entrance time and the value of the payment flag 11009 in the store data (FIG. 11) of the store is 0 (prepaid), the usage time is The time obtained by adding the required time 11012 of the store is set as the true departure time ((a) 12003), and if the value of the payment flag 11009 is 1 (postpay), the use time is set as the true departure time (FIG. 12 (b) 12005). From the entrance to the true departure time, the residence time at the entrance station (FIG. 12 (a) 12004, FIG. 12 (b) 12007).
(A) When there is a usage record at the participating station: If the value of the payment flag 11009 of the store data (FIG. 11) of the store is 1 (postpay) instead of the entry time, the usage time is The time obtained by subtracting the required time 11012 of the store is set as the true arrival time, and if the value of the payment flag 11009 is 0 (prepaid), the use time is set as the true arrival time, and the stay from the true arrival time to the entry is retained at the participating station. Time.
(C) When there is a usage record at a midway station (transfer station): The time spent in the station yard without going out of the ticket gate at the midway station is calculated in the same way. If the value of the payment flag 11009 in the store data of the store (FIG. 11) is 1 (post-payment) with respect to the earliest use time in the use record of the store (12008 to 12010) in the same station, the store is determined from the use time. The time obtained by subtracting the required time 11012 is set as the true arrival time (FIG. 12 (c) 12011), and if the value of the payment flag 11009 is 0 (prepaid), the use time is set as the true arrival time, If the value of the payment flag 11009 of the store data (FIG. 11) of the store is 0 (prepaid) with respect to the earliest use time in the use record, the time obtained by adding the required time 11012 of the store to the use time is true If the value of the payment flag 11009 is 1 (postpay), the use time is set as the true end time. The residence time at this station is defined as the time from the true start time to the true end time (FIG. 12 (c) 12013).

A similar approach is used to estimate the residence time around the station.
(A) When there is a usage record after participation: After the participation, the latest usage time in the usage record around the participating station is used, and the value of the payment flag 11009 of the store data (FIG. 11) of the store (13002) is If 0 (prepaid), the time obtained by adding the required time 11012 of the store to the use time is set as the true end time (FIG. 13 (a) 13003). If the value of the payment flag 11009 is 1 (postpay), the use time is set to true. The end time (FIG. 13 (b) 13005) and the true end time from the entry is the residence time around the entry station (FIG. 13 (a) 13004, FIG. 13 (b) 13007).
(A) When there is a usage record before entering: The value of the payment flag 11009 in the store data (FIG. 11) of the store (13006) using the earliest use time among the usage records around the entrance station before entering. If the value is 1 (postpay), the time obtained by subtracting the required time 11012 of the store from the use time is set as the true start time. If the value of the payment flag 11009 is 0 (prepaid), the use time is set as the true start time. The time from the start time to the admission is the residence time around the entrance station.

  The estimation of the stay time described above (described with reference to 703 in FIG. 7, FIG. 8, and FIG. 9) is a method for estimating the total stay time at the entrance station or the exit station. This estimated time is Ts, and further, according to the above calculation, for the entrance station (E station), the departure station (X station), and the transfer station (C station), within the station premises (within the ticket gate) and around the station (outside the ticket gate) The estimation result of the residence time was obtained. The residence time Tei in the entrance station, the residence time Teo around the entrance station, the residence time Txi in the entrance station, the residence time Txo around the entry station, the residence time Tci in the transfer station, and the residence time Tco around the transfer station Then, the stay time is corrected according to the following rules. Further, Tr is the station residence time (= Ts−Tei−Txi−Tci, except 0), excluding the correction value.

Admission station (E station) stay time Tse:
(1) When E station is Navabari 4 station and X station, C station is more frequently used (Fig. 4, use frequency 413): Tse = Tei + Teo + Tr
(2) When E station is less used than Nawa Bali 4 station and either X station or C station (or both): Tse = Tei + Teo
(3) E station other than Nawa Bali 4 station: Tse = Tei + Teo
Residence time of station (X station) Tsx:
(1) When station X is used more frequently than Nawabari 4 station, E station and C station: Tsx = Txi + Txo + Tr
(2) When X station is less used than Nawabari 4 station and E station or C station (or both): Tsx = Txi + Txo
(3) X station is not Nawabari 4 station: Tsx = Txi + Txo
Transit station (C station) stay time Tsc:
(1) When station C is used more frequently than Nawabari 4 station, E station, and X station: Tsc = Tci + Tco + Tr
(2) When station C is less frequent than Nawabari 4 station and either E station or X station (or both): Tsc = Tci + Tco
(3) Station C is not Nawa Bali 4 station: Tsc = Tci + Tco
When e, x, and c are Nawa Bali 4 stations as described above, the calculated stay time Tsi (i = {e, x, c}) is added to the sum of stay times of the corresponding station index values (414 in FIG. 4). , The square value of Tsi is added to the sum of squared stay times (415 in FIG. 4). When the station is not Nawabari 4 station, in addition to the sum of stay times of index values of other stations (414 in FIG. 4), the square value of Tsi is added to the sum of squares of stay time (415 in FIG. 4).

  When changing to another company's route, or when getting off and shopping, the ticket gate passes, so the same processing as above is performed by dividing into a plurality of entrance-exit sections.

  In order to improve accuracy, use the transportation history such as buses after purchasing behavior around the station and the usage history of stores in the city that are not around the station. Can be estimated. In addition, in order to eliminate the possibility of purchase behavior (payment with cash) that does not remain as a record, the last purchase history that remains as a record is a product that rarely makes another purchase after purchase, such as sugar beet, sweets, or fresh flowers It is also possible to use knowledge such as whether or not.

  Furthermore, the model of FIG. 9 is established when the train is operated on time, but if the operation changes greatly (delays) due to an accident or the like, the calculation is performed taking into account the delay information, This is dealt with by not performing the calculation.

  Thus, the process of 703 in FIG. 7 is completed. Subsequently, the process of 704 in FIG. 7 is performed. Here, a simple index value for card writing shown in FIG. 5 is calculated.

  The same information part as the index value shown in FIG. 4, that is, 401 is copied to 501, 402 is copied to 502, and each station code 411 and station flag 412 of Nawa Bali 4 station are copied to 511 to 518. The number of uses 519 to the convenience use amount 524 are values obtained by taking the sum of 413, 414, 416, 417, 418, and 419 for Nawa Bali 4 stations in FIG.

  This completes the description of FIG. Hereinafter, the description returns to the calculation server 1082. The user classification index file stores the index value master shown in FIG. 4, the simple index value for the card shown in FIG. 5, and the user personal setting file shown in FIG.

  The partner store advertisement file stores the store data shown in FIG. 11 and the advertisement data shown in FIG. The advertisement data shown in FIG. 14 includes a company that advertises, a store (affiliation), store codes (14101 to 14003), and a period (14004 and 14005) where the advertisement is issued. Furthermore, as information that limits the target to be advertised, the station flag (14006) (the station flag is FIG. 412), the lower limit number of stations (14007), the target tendency (14008, convenience or entertainment), the target The user participation frequency (14009) and the use time zone (14010) are held. The message (14011) is specific contents of the advertisement.

  Next, the function of the partner information distribution server 1083 will be described.

  The partner information distribution server 1083 edits information for the partner in accordance with the index value of the user calculated by the calculation server 1082 for the partner company. The partner information distribution server 1083 edits two pieces of information. One is information that a partner company sales promotion / planner (1331 in FIG. 1) refers to in order to know the macro information of the user. This information is referred to by the terminal device of the partner company (1321 in FIG. 1). This is called analysis information for partners. The other is when the customer is in the store (actually at the time of payment), the salesperson who receives the customer uses the POS cash register terminal for payment (1033, 1035 in FIG. 1) for the sales promotion tailored to the user. Information. This is called sales promotion information for partners.

  First, analysis information for partners will be described. FIG. 15 is an example of the screen. The entire screen 15001 is roughly divided into upper (15003, 15004) and lower (15005, 15007, 15009) information. The upper information is information for each store. First, a sales promotion / planner in a partner company selects a store A to be investigated from a store selection pull-down menu 15002. The selection information is transmitted to the partner information distribution server, and the partner information distribution server that has received the information searches the store data (FIG. 11) for the station code S where the store is located corresponding to the store code of the store A. Ask.

  Next, among the index values of all users (FIG. 4), the number of uses (413) and the number of uses (423 to 423) of users whose one of the four stations of Nawabari is the station of the station code S. 440) for each item. Then, values of 18 items obtained by dividing the total value of 423 to 440 by the total value of the number of uses (413) are obtained, and the information is transmitted to the terminal device 1321 and displayed on the graph 15003. This represents a utilization rate for each day of the week and time of a user who uses the station S as a narrow bar.

  A graph 15004 displays values obtained by tabulating the usage history records of FIG. Here, regarding the three types of total values, the total values for 18 cases in the day of the week / time zone are obtained in the same manner as in the case of 15003. The first aggregated value is the sum total of the users of the station S every day, the second aggregated value is the sum total of the users who have a purchase record at the store A, and the third aggregated value is the record of purchases at the store A. It is the sum total of the users who use the station as Nawa Bali. With respect to the third total value, it is determined with reference to 102 to 105 in FIG. This tabulation indicates how many of the station users use the store A, and how many of those users use the station as a narrow bar.

  In the lower row, regarding the two periods specified in the fields 15006 and 15008, the number of users per day who record purchases at store A during that period, the amount paid per day, and the entire store day The sales per unit, the number of people who set the station S as Nawabari, the payment amount of those who are Nawabari, and the sales per day are tabulated using history records as in the case of 15004. 1507 is displayed. A difference between these items is displayed in 15009.

  Next, sales promotion information for partners will be described. FIG. 16 shows an example of a screen of a POS cash register terminal or the like (FIG. 1 1033, 1035) for a salesperson who serves as a customer. An area 16002 in the upper part of the screen 16001 displays a sales amount, a received amount, a change amount, and the like, which are functions as an original POS cash register terminal. An area 16003 at the bottom of the screen is sales promotion information for a partner referred to by a salesperson. This part is composed of an area 16004 that represents the attributes of the user on the left side and an area 16005 that displays the advertisement message on the right side. The 16003 portion of the screen shows the index value of the user by referring to the user classification index from the ID when the facing user presents the contactless IC card for settlement and reads the information. Find and change the content based on that value. The part 16004 is an indicator corresponding to the index value of the user who meets, whether or not the station where the store is located is included in the Nawa Bali 4 station, and the tendency of the store that the store uses frequently (recreational or not) It is expressed in a bar graph format according to whether it matches the convenience) or whether the station is used frequently or not. In other words, it is one of Nawa Bali Stations, and the value is high when the tendency of frequently used stores matches and the station is frequently used. Further, referring to the advertisement data (FIG. 14), the advertisement of the store (14001 to 14003 matches), the period matches (the day is between the start date and time 14004 and the end date and time 14005), and the target station flag 14006 is It matches the tendency of the station, the number of times the user uses the corresponding station is higher than the target station number 14007, the store tendency 14008 matches, the user uses the station more frequently than the target frequency 14009, and the target time 14010 If the user's stay time is longer than that, one of the advertisement data satisfying the conditions of 14006 to 14010 is selected at random, or if there is no data that satisfies all, one of the advertisement data having the most conditions to be satisfied Is selected at random, and the message 14011 of the selected data is displayed in the advertisement message 16005.

  Next, functions of the user information server 1084 will be described. The user information server 1084 edits information for the user according to the index value of the user calculated by the calculation server 1082 for the user. The user information server 1084 has two major functions and edits four types of information. The user web edits three types of data.

  One is user-provided information for the user himself / herself to check his / her Nawa Bali station usage status and to refer to his / her own advertisement. This is common with information (FIG. 17) for reference from a Web browser of a terminal device such as a personal computer through the Internet and information (FIG. 18) for reference from a browser function of a mobile phone using the mobile phone network. Edit two pieces of information based on information. The other is information displayed on the screen for the user of the POS cash register terminal when paying at the store, or displayed on the screen for the user when passing through the ticket gate. This is called advertisement information for users (FIG. 19). The last one is that this advertising information for users is transmitted to the user's mobile phone 1432 at a preset time or when passing through the ticket gate by the mail function of the mobile phone using the information distribution function 1412. It is.

  Here, with Web commerce and mobile phones, recommendations can be presented when the user is operating the terminal, so there is a possibility that it can induce “buy” behavior at the time of purchase, whereas non-contact IC cards The user cannot specify the user until the moment of settlement, and the card itself does not have a function for displaying information online, so the timing at which information can be presented is limited. Therefore, in the present embodiment, a non-contact type IC card is used by a ticket gate or a ticket machine for the purpose of providing information to a user of a non-contact type IC card whose information presentation timing is limited at an appropriate timing. It is also characterized by displaying information on the screen for the user when using. Thereby, before using a station store (not after purchase), advertising information for users can be provided to a user.

  First, information provided for users will be described. FIG. 17 shows an example of the screen. The entire screen 17001 is roughly divided into seven areas. A nickname preset by the user is displayed in the title portion 17002 on the upper left. On the upper right, the date of the day and the expected weather (prediction of weather, temperature, etc.) of the day are displayed (17003). In the status information 17004 at the upper left of the screen, the index values 413, 414, and 416 related to the user's index value (FIG. 4) 402 and Nawa Bali 4 stations are displayed in a band graph format. Convenience book information 17005 displays station timetables of Nawabari 4 stations and links to accidents and delay information at that time on railway lines passing through those stations. The today's recommendation column 17006 in the center is a store having a store code (11005) that matches the date 14003 from the advertisement data (FIG. 14) and that is between the start date and time and the user's Nawa Bali 4 station. For one record in the data (FIG. 11), the station code 11008 matches one of the four stations of Nawabari), the target station flag 14006 matches the user's Nawabari station, and the store trend 14008 matches. If the user's station usage frequency is higher than the target frequency 14009 and the user's stay time is longer than the target time 14010, the advertisement data satisfying the conditions from 14006 to 14010 are listed in order, and the most satisfying condition The advertisement data is displayed in descending order. Furthermore, it can also be added on condition that the number of corresponding stations of the user is higher than the target station number of 14007.

  The news column 17007 is the news published on the Web by the station, the route that passes through the station, or the station premises and the surrounding stores (the company that manages the station), with the new posting time up, Enumerate links.

  The banner list column 17008 is a list of banners linked to the homepage of a store (or a company that manages the store) at the user's Nawa Bali 4 station. The store is located at Nawa Bali Station, which is frequently used, and stores that tend to be used frequently by users at the station are placed at the top.

  The screen 18001 in FIG. 18 has the same information as the screen 17001 in FIG. For mobile phones, reduce the amount of information that can be displayed. 18002 is a title, and 18003 is a status display. One item is selected at random from the station usage type 402 of the user's index value (FIG. 4) and the indices 413, 414, and 416 related to the four stations of Nawa Bali. The index values such as the frequency for each station are expressed in shades (red is large and blue is small). Convenience book 18004, recommendation 18005, and news 18006 report the number of pieces of information to be displayed in three.

  The advertisement information for users in FIG. 19 is displayed on a display device that is visible to the user at the time of payment at a POS cash register terminal or the like. The entire screen 19001 consists of two areas. When the station is included in the user's Nawabari 4 station, the status area 19002 graphically displays four values of index values 413, 414, 416, and 417 relating to the station as shades. The message area 19003 is from the advertisement data (FIG. 14), and the store date (FIG. 14) has a store code (11005) that matches the start date and end date and stores at the user's Nawabari 4 station (14003). 11) and the station code 11008 matches any of the four stations of Nawabari), the target station flag 14006 matches the user's Nawabari station, the store trend 14008 matches, and the target When the frequency of use of the station by the user is higher than the frequency 14009 and the staying time of the user is longer than the target time 14010, the advertisement data that satisfies the conditions of 14006 to 14010 are listed in order, and the advertisement is further performed in the order of the most satisfying conditions. The data is displayed. Furthermore, it can also be added on condition that the number of corresponding stations of the user is higher than the target station number of 14007. As described above, the advertisement data satisfying a predetermined condition is displayed to the user by comparing the three indexes of the station use frequency, the store use tendency, and the station stay time with the index indicated by the pre-recorded advertisement data. ·provide.

  Here, the reason why the user's behavior tendency is analyzed based on the three axes of the station use frequency, the store use tendency, and the station stay time will be described. Considering the purpose of identifying the purpose of a user's use of a station (and the surrounding commercial facilities) from the person's behavior, first of all, how often the station itself gets on and off the station It can be determined whether the user is an active station user or not, depending on how long he / she stays in the station. The frequency of use can be interpreted as having a different purpose between frequently using only a specific station or a plurality of fixed stations and using many stations unspecified.

  Furthermore, the purpose of use can be further discriminated depending on at which station there are many purchases and what kind of tendency to purchase there. In particular, with regard to purchasing, since there are various industries and business types, it is first determined what kind of store to use rather than looking at individual stores and individual products. At the core of this is the use of highly convenient stores such as mainly purchasing foods (e.g. side dishes) and daily necessities necessary for daily life, or mainly using fast food that can be eaten alone in a short time. The purpose can be discriminated naturally when using high-entertainment stores such as fashion items such as fashion and restaurants where people eat while talking.

  In addition, when a store with high convenience is generally used, it is considered that one station stay time is relatively short. Therefore, even if there are many stores that are convenient to use, if the staying time is long, after enjoying window shopping without actually purchasing, there is a tendency to purchase everyday items and go home Can be analyzed. Thus, a composite tendency can be discriminated by combining the three axes. In the non-contact IC card usage history data, it is possible to extract information necessary for the above discrimination from this station because it is recorded which station is used at what time and which store is used. is there.

  The information sent by the information distribution selects an advertisement message (14011 in FIG. 14) regarding the four stations of Nawabari Station based on the same selection criteria as the information for the advertisement information terminal for users. Then, it is sent to the user's mobile phone at a predetermined time or at a timing when it passes the ticket gate.

  In the above-mentioned editing of information for users, the station usage style that the user himself / herself has input in advance, such as “Guest of business trip”, “Guru of sale”, “Guru of delicatessen”, etc. For example, “Guests of business trips” often use during the day, use more frequently at stations other than Nawabari 4 stations, and spend some time at transfer stations, but compare the time per stay. It is possible to provide a game-like service, such as providing a standard for each style, such as giving a title as “expert” when all the corresponding items of the user's index value exceed a predetermined value.

  In the above example, the usage record of the traffic contactless IC card is used to obtain the index value. In addition to this, payment by credit card, other IC card type electronic money is used. In addition, the use history such as location information acquired by a mobile phone or the like can be used to improve the accuracy of index values such as station stay time and payment amount at a store.

  According to the present invention, the index value shown in FIG. 5 is recorded on the non-contact IC card held by the user. This information can be recommended in accordance with the user's behavior tendency by reading the index value of the card user even in a reader / writer device not connected to the network 1071 in FIG.

  From the above, by characterizing the behavior in the station premises and around the station, which is the basis of the present invention, by the frequency of stay, the time, and the usage tendency of the station store (requiring entertainment or convenience), each use It is possible to recommend stores and products that match the style of use at the station.

1011, 1012, 1013, 1014, 1015 User 1021, 1022, 1023, 1024, 1025 Non-contact IC card 1031 Ticket gate 1032 Ticket vending machine 1033 Ticket vending machine (ticketing terminal)
1034 Information terminal 1035 POS cash register terminal 1111, 1112, 1113, 1114, 1115 CPU
1121, 1122, 1123, 1124, 1125 Non-volatile memory 1131, 1132, 1133, 1134, 1135 Radio module 1141, 1142, 1143, 1144, 1145 Antenna 1041, 1042, 1043, 1044, 1045 Reader / writer device 1211, 1212, 1213 1214, 1215 CPU
1221, 1222, 1223, 1224, 1225 Memory 1231, 1232, 1233, 1234, 1235 Radio module 1241, 1242, 1243, 1244, 1245 Antenna 1051, 1052, 1053, 1054, 1055, 1063, 1065 Display device 1071 Network 1082 Calculation Server 1311 Network interface 1312 Processor 1313 Memory 1314 Recording device 1083 Partner information distribution server 1084 User information server 1411 User Web function 1412 Information distribution function 1072 Internet 1321 Partner person terminal 1331 Partner person 1431 Personal computer 1432 Mobile phone 1441 1442 User 1073 Mobile phone carrier 201 1 Notification 301 user ID enumeration 202 index value recalculated 203 index value updating 204 the index value updating partial history
302 Date and time 303 Station store code 304 Terminal code 305 Type 306 Amount 307 Balance 308 Calculation standard 1
309 Calculation standard 2
401 Start date 402 Station use type 403 Station use station code 411 Station code 412 Station flag 413 Number of uses 414 Total stay time 415 Total sum of stay time 416 Recreation store 417 Convenience store number 418 Recreational use amount 419 Convenience Usage amount 420 Recreational usage amount squared 421 Convenience usage amount squared sum 423 Weekday morning usage 424 Weekday morning usage 425 Weekday afternoon usage 426 Weekday afternoon usage 427 Weekday evening usage 428 Weekday evening usage Number of use 429 Number of times used in the morning in the weekend 430 Number of times used in the morning in the weekend 431 Number of times used in the afternoon in the weekend 432 Number of times used in the afternoon in the weekend 433 Number of times used in the evening in the weekend 435 Number of times used in the evening on the weekend 435 437 Number of public holidays in the afternoon 438 Number of public holidays in the afternoon 43 Holiday evening of the number of times of use 440 holiday night use the number 501 start date 502 Station use species number 511 Station code 1
512 Station flag 1
513 Station code 2
514 Station flag 2
515 Station code 3
516 Station flag 3
517 Station code 4
518 Station flag 4
519 Number of times of use 520 Total time of stay 521 Number of recreational stores 522 Number of convenient stores 523 Revenue use amount 524 Convenience use amount 601 User ID
602 Balance 603 Station code 701 Station estimation 702 Trend calculation 703 Stay time estimation 704 Card index calculation 801 Station A entry time 802 Station B entry time 803 Elapsed time 804 Train travel time 805 Walking travel + waiting time 806 Walking time 807 Residence time 808 Station B store X use time 809 Station C stay time 810 Station A store Y use time 811 Station A vicinity residence time 812 Store Y use time 813 Station B store Z use time 814 Station B vicinity Residence time 815 Vending machine usage time at station C 901 Time required 902 Number of users 903 Average time required 904 Minimum time required 905 Station residence time 906 Typical value 907 Minimum time required for user 908 Value obtained by adding time deviation to typical value 909 Station residence time 911 time deviation (difference between typical value and shortest value)
912 Time deviation 913 Station B entry time 914 Station residence time 101 User ID
102 Station code 1
103 Station code 2
104 Station code 3
105 Station code 4
106 Station code A
107 Station code B
109 Typical Time 110 Time Deviation 111 1 Record 11001 Company Code 11002 Company Name 11003 Store Code 11003 Store Name 11005 Store Code 11006 Store Name 11007 Industry Code 11008 Station Code 11009 Payment Flag 11010 Internal / External Flag 11011 Trend Value 11012 Required Time 11013 Average Time 11014 11015 Closed time 11020 Number of terminals 11021-11023 Terminal code 11030 Ticket gate code 11031 Time required 12001 Station A entrance time 12002 Store B use time 12003 True departure time 12004 Residence time 12005 Store C use time 12006 Store C use start time 12007 Residence time 12008 Store F use time 12009 Store G use time 12010 Store H interest Time 12011 True start time 12012 True end time 12013 Residence time 13001 Station A entry time 13002 Store B use time 13003 True end time 13004 Residence time 13005 Store C use time 13006 Store C use start time 13007 Residence time 14001 Company code 14002 Store code 14003 Store code 14004 Start date and time 14005 End date and time 14006 Target station flag 14007 Target station number 14008 Target trend 14009 Target frequency 14010 Target time 14011 Message 15001 Partner analysis information screen 15002 Store selection menu 15003 Store S Usage frequency graph 15004 Station S user and store usage performance graph 15005, 15007 Number of users per day, support Amount of money, graph of all store sales 15006, 15008 Period input field 150009 15005 and 15007 Difference graph 16001 Promotion information screen for partners 16002 POS cash register function screen 16003 Promotion information for partners 16004 User attribute display section 16005 Advertising message section 17001 For Web Provided information screen for users 17002 Page title 17003 Basic information area 17004 Status area 17005 Convenience book area 17006 Recommended area today 17007 News area 17008 Partner company banner advertisement 18001 Provided information screen for mobile phone users 18002 Page title 18003 Status Area 18004 Convenience book area 18005 Recommended area 18006 News area 19001 Advertising information screen for users 19002 Status area 19003 Message area

Claims (10)

Configured to include a transport system settlement means each user to the mobile, and the reader-writer device installed in commercial facilities around stations and station, the behavior trend analysis server connected via a network to the reader-writer device A behavior analysis system,
The transportation settlement means has a memory for holding the use history of the transportation settlement means including the entry / exit history of the station and the purchase history at the commercial facility to a limited number of the latest,
The behavior trend analysis server
A network interface connected to the network;
A processor connected to the network interface;
A recording device connected to the processor,
The processor
The user classification index of the recording device is calculated by receiving the usage history read from the memory of the transportation settlement means via the network interface, calculating the number of station usages based on the entry / exit history included in the usage history. Stored in a file
Based on the purchase history included in the usage history and the store data stored in the recording device, store usage tendency indicating convenience or entertainment is calculated and stored in the user classification index file,
Based on the entry / exit history included in the usage history, the station stay time is calculated and stored in the user classification index file,
The advertisement data satisfying a predetermined condition is selected by comparing the store use frequency, the store use tendency, and the station stay time with the store advertisement data recorded in advance in the recording device, and the selected advertisement data. A behavioral trend analysis system characterized by output. The behavior tendency analysis system according to claim 1,
The processor, from the usage history, calculates the number of stations that the user of the public transport payment means using, and selects the predetermined condition is satisfied advertisement data based on the number of the station Behavior trend analysis system . The behavior tendency analysis system according to claim 1 or 2,
The processor, for each station of the station and the other to be specified in advance by the user of the transportation system settlement means, actions and calculates the station the number of times of use, the store usage trends, and the above-mentioned station stay time Trend analysis system . The behavior trend analysis system according to any one of claims 1 to 3,
The processor transmits the station usage count, the store usage tendency, and the station stay time to the reader / writer device,
When the traffic settlement means communicates with the reader / writer device, the reader / writer device records the number of times the station is used, the store use tendency, and the station stay time in a memory in the traffic settlement means. Behavior trend analysis system . The behavior tendency analysis system according to any one of claims 1 to 4,
The said processor corrects the said station stay time using the purchase log | history contained in the said use log | history, The behavior tendency analysis system characterized by the above-mentioned. The behavior tendency analysis system according to any one of claims 1 to 5,
The processor transmits the selected advertisement data to the reader / writer device,
A behavior trend analysis system in which the selected advertisement data is displayed on a display unit connected to the reader / writer device when the traffic settlement means communicates with the reader / writer device. Behavioral tendencies that are connected via a network to a traffic payment means carried by each user, a reader / writer device installed in each station and a commercial facility around the station, and a reader / writer device that communicates with the traffic payment means A behavior trend analysis method executed in a behavior trend analysis system including an analysis server,
The entry / exit history of the station and the usage history of the transportation settlement means including the purchase history at the commercial facility are written into the transportation settlement means from the corresponding reader / writer device, and the transportation settlement means is the most recent constant A step of keeping the usage history in the internal memory only for the number,
The behavior trend analysis server receives the use history read from the memory of the traffic payment means , calculates the station use frequency based on the entrance / exit history included in the use history, and prepares for the behavior trend analysis server Storing in a user classification index file of the recording device;
The behavior trend analysis server calculates a store usage trend indicating convenience or entertainment based on the purchase history included in the usage history and the store data stored in the recording device, and stores it in the user classification index file. Storing, and
The behavior trend analysis server calculating a station stay time based on an entry / exit history included in the usage history and storing the station stay time in the user classification index file;
The behavior trend analysis server selects the advertisement data satisfying a predetermined condition by comparing the store use frequency, the store use tendency, and the station stay time with the store advertisement data recorded in advance. And a step of outputting the advertisement data. The behavior tendency analysis method according to claim 7,
The above behavior trend analysis server, a feature for each station of the station and the other to be specified in advance by the user of the transportation system settlement means, the station number of times of use, the store usage trends, and to calculate the station dwell time A behavior trend analysis method. The behavior tendency analysis method according to claim 7 or 8,
The behavior trend analysis server has a step of transmitting the station use count, the store use tendency, and the station stay time to the reader / writer device,
When the traffic settlement means communicates with the reader / writer device, the reader / writer device causes the station usage count, the store usage tendency, and the station stay time to be stored in the memory within the traffic settlement means. Behavior trend analysis method to be recorded. The behavior tendency analysis method according to any one of claims 7 to 9,
The behavior trend analysis server has a step of correcting the station stay time using a purchase history included in the usage history.

Images