JP2010244433A - Information processing system and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To propose a mechanism for indirectly inferring facilities on which a user drops or which the user stays while using history information which is intermittently obtained according to user's action or movement. <P>SOLUTION: An operator inputs a place or an event to a system as a retrieval condition. The system extracts the history information of each mobile object which is collected by a location information acquisition apparatus arranged in the periphery of the specified place from the intermittently acquired history information. Then, the system verifies the extracted history information with at least one of the extracted history information. stay time and past action history of each mobile object, a distance from the place input by the operator, and time that the even has been held and infers a mobile object having high possibility of corresponding to the retrieval condition. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an information processing technique for extracting meaningful information from movement history data collected intermittently with the movement and behavior of a moving body. For example, the present invention relates to a technique for indirectly estimating individual or collective behavior or behavior (for example, hobbies, interests, interests) of a moving object.

  Due to the miniaturization and high performance of wireless devices and computers, a method for easily and accurately acquiring the position information of a moving object (including humans, animals, and objects) has been put into practical use.

  The method for acquiring the position information of a moving body includes a method using radio waves from an artificial satellite such as GPS (Global Positioning System), a self-contained navigation using a sensor such as a gyro sensor, and a mobile phone. There are various methods such as a method of specifying the approximate position of a mobile phone or PHS using a base station of PHS (Personal Handyphone System).

  Today, location histories acquired using these methods are used for marketing, infrastructure development, urban planning, etc. through collection, aggregation, analysis, analysis, and the like. For example, the result of analyzing or analyzing the movement tendency of the entire moving body and the behavior of each moving body is used for marketing, infrastructure development, city planning, and the like.

  For example, a method (mobile marketing method) for distributing marketing information determined according to an action log (including a terminal position and its time) and user attributes has been proposed for a user possessing a portable terminal (patent) Reference 1). In addition, a method has been proposed in which a history of the visited place is collected from the position information of the moving body, and trends such as the preference and visiting pattern of the moving body are extracted (see Patent Document 2). Further, a method for estimating a user's current destination candidate based on the user's past behavior is disclosed (see Patent Document 3).

  On the other hand, railway IC (Integrated Circuit) tickets, which have recently been introduced, provide convenience to passengers, such as not requiring the hassle of buying tickets and can also be used as electronic money. It produces the effect of improving efficiency. By the way, an ID (Identification) code for identifying each IC card (including an IC chip built in a portable terminal) used as a boarding ticket is given. Therefore, if the ID of the ticket and the location of the ticket gate through which the ticket passes can be associated, it can be regarded as a kind of position information acquisition device. The same can be said for the toll road payment system using ETC (Electronic Toll Collection System). Note that these are not limited to transportation systems such as railways and roads. If purchase information using electronic money or point cards can also be associated with the location information of the store where the purchase was made, similarly, Can be regarded as one.

  Note that a mechanism (for example, “GOPASS”) that distributes marketing information in accordance with the position information of the ticket gate that the user has passed using an IC ticket has already been put into practical use (see Non-Patent Document 1).

Patent Gazette: Patent No. 3726748 Published patent publication: JP 2000-155757 Published patent publication: JP 2000-20548

"Information distribution system" goopas "based on location, time and behavior": OMRON TECHNICS Vol.43 No.3 (Vol.147), pp287-292, 2003

  By the way, there are many various methods that are currently put to practical use, which basically estimate the direct destination and staying place of a moving body based on the acquired movement history and position history. In other words, the current method is based on location information directly acquired from stores, restaurants, landmarks, and other various facilities (so-called POI: Point Of Interest) where a certain moving body exists or stays. Guess history, interests, hobbies, etc.

  For example, in the invention of Patent Document 1, information that associates a semantic location called a “location information database” with an actual spatial position is compared with a user's location history obtained by GPS to determine the user's actual location. Identify. And the method of delivering the content according to the specified place to a user is provided. Further, for example, the invention of Patent Document 2 presumes the visited place based on the position information obtained by the position information acquisition means such as GPS and the situation data related to the moving situation, and specifies the visited place through a question to the user. Provide processing functions. Further, for example, the invention of Patent Document 3 determines that the user has stopped at the destination when the position information by GPS remains in a shop for a certain time or more.

  All of these methods are based on the premise that the movement history of the moving object is continuously and accurately acquired and simultaneously recorded. That is, it is necessary to acquire pinpoint position information in time series from each moving object. However, automatically acquiring pinpoint position information without the consent of the moving body is considered to have a high psychological burden on the user from the viewpoint of privacy and the like. On the other hand, the method of collecting location information through connection to a site, a user's specific operation with respect to an application, etc. can reduce the psychological burden imposed on the user. However, in this method, since the user is required to provide the location information, there is a possibility that data collection may be leaked, for example, the information cannot be acquired even though the user is actually staying at a certain facility.

  On the other hand, in the case of history information collected when passing through a ticket gate of a station or a toll gate on a toll road, the position history of passing points is reliably recorded while reducing the psychological burden on the user. be able to. However, stations and toll booths generally do not coincide with final destinations and are merely passing points. For this reason, it cannot prove that it existed in a certain facility by the history information (user ID, passage date and time information, and position information) indicating that it passed through these gates.

  Therefore, the inventor proposes the following mechanism. First, an operator inputs a place or event as a search condition into the information processing system. In the information processing system, it is collected by the position information acquisition device arranged around the place specified as the search condition (including the place where the event was held) from the history information acquired intermittently Only the history information for each moving object is extracted. Next, the information processing system uses the extracted history information, the extracted history information for each moving object, the staying time, the past behavior history, the distance from the place entered by the operator, or the time when the event was held The mobile object that is highly likely to have visited the place entered as the search condition is indirectly estimated.

  The present invention is not limited to the information processing system described above, and can be configured as an information processing method. The present invention can also be realized by a computer program that realizes the information processing system described above, a recording medium that records the program, and other various aspects.

  Incidentally, when the present invention is configured as a computer program or a recording medium recording the program, the program according to the present invention may be configured as a whole program recorded on the recording medium, or the program according to the present invention may be You may comprise as a part of program recorded on a recording medium. As a recording medium used for distribution, various volatile storage media and nonvolatile storage media that can be read by a computer such as a flexible disk, a CD-ROM, and a DVD-ROM can be used.

  According to the present invention, while reducing the psychological burden given to the user, based on historical information on the position of passing points that can be reliably collected, the user visited a place designated as a search condition or participated in an event held at the place. Individual or collective mobiles can be estimated.

It is a figure showing the example of whole composition of the information processing system concerning a form example. It is a figure explaining the positional relationship of a facility and a position specific terminal. It is a figure which shows the example of a data structure. It is a figure which shows the example of a whole process sequence. It is a figure explaining the example of an execution procedure of collation and a user extraction process. It is a figure explaining the weighting example of likelihood calculation. It is a figure which shows the example of a user's action shown simplified. It is a figure which shows the example of data. It is a figure which shows the example of data. It is a figure explaining the example of extraction operation of the user who participated in the exhibition. It is a figure which shows the example of a screen (for event selection) of the terminal for control. It is a figure which shows the example of a screen (for extraction result display) of the terminal for control. It is a figure which shows the example of a screen (example which pays attention to a user) of the terminal for control. It is a figure which shows the example of a screen (example which pays attention to a user) of the terminal for control. It is a figure explaining the transition relationship between screens. It is a figure which shows the other example of a user's behavior shown simplified.

Hereinafter, based on the drawings, exemplary embodiments according to the present invention will be described in order.
(1) Example 1
A first embodiment of the information processing system will be described with reference to FIGS.
(1-1) System Configuration and Function FIG. 1 shows an overall configuration example of an information processing system (hereinafter simply referred to as “system”) according to Embodiment 1. The system includes a server (001) that executes data storage and processing, a location specifying terminal (006), and a control terminal (008). These terminals are connected to each other via a network (004). Further, the position specifying terminal (006) in the form example refers to any device that can acquire the position of the mobile object, and corresponds to, for example, a gate of a transportation facility or a base station of a communication device. In addition, the mobile body in this embodiment is a person, animal, or object (for example, a car equipped with ETC) that possesses or carries an ID notification terminal (for example, card) (005) that can identify a user through electronic communication. . In the following description, it is assumed that the moving body is a person and possesses a card-shaped ID notification terminal (005).

  When the user (002) passes the vicinity of the location specifying terminal (006) while holding the ID notification terminal (005), the location information including the user ID for identifying the user (002) and the passage date is stored on the server (001). (301). This location information transmission (301) may be executed simultaneously with the acquisition of the location information, or may be executed at an appropriate timing after the location information is temporarily stored in the data storage unit (009) of the location specifying terminal (006). May be. Further, depending on the system configuration, an ID notification terminal (005) possessed by the user (002) acquires position information from the position specifying terminal (006) and transmits the acquired position information to the server (001). May be.

  The system according to this embodiment can be applied to a railway boarding system, for example. In this case, the IC boarding ticket corresponds to the ID notification terminal (005), and the ticket gate corresponds to the position specifying terminal (006). Further, the system according to this embodiment can be applied to a passage management system such as a toll road. In this case, the ETC card corresponds to the ID notification terminal (005), and the entrance / exit gate corresponds to the location specifying terminal (006). The system according to this embodiment can also be applied to a payment system including, for example, a terminal or card that holds a user ID as electronic information and a terminal that holds position information. In this case, when communication or settlement is realized between a card or credit card storing electronic money and a merchandise terminal such as a settlement terminal or vending machine, position information including a user ID and a communication date and time can be collected. Is used. At this time, a settlement card or the like corresponds to the ID notification terminal (005), and a settlement or product sales terminal corresponds to the position specifying terminal (006).

  Generally, the user (002) moves to a facility (007) such as the nearest stadium or exhibition hall through a location terminal (006) such as a ticket gate or an entrance gate. That is, it participates in events such as sports matches and exhibitions (304). After the end of the event, it passes again from the facility (007) near the nearest location terminal (006) and returns to the home or work. Here, the location specifying terminal (006) passed before the start of the event and the location specifying terminal (006) passed on the return are not necessarily the same. For example, when there is a facility (007) between stations or when there are a plurality of ticket gates for one station, the travel route becomes more arbitrary, and the travel route does not match between going and returning.

  The server (001) is mainly composed of two parts, a data storage unit (100) and a control unit (200). In the data storage unit (100), location log data with ID (101) for storing the location history of each user sent by location information transmission (301), and user data (102) for managing information about the user. , Extraction result data (106), collation data (110), and the like. The collation data (110) is performed at the location specifying terminal data (103) for managing information on each location specifying terminal (006), facility related data (104) for managing information on the facility (007), and the facility. Event related data (105) for managing information related to the event is included.

  The control unit (200) includes a location log-event collation processing unit (201) for associating a user's location log with event participation, a collation data maintenance processing unit (202), a data communication / input / output processing unit (203), Including a database access processing unit (204).

  When an operator (003) who performs data analysis inputs (selects) conditions to the server (001) through the control terminal (008) (302), a predetermined collation process is performed by the position log-event collation processing unit (201). And the result can be obtained (303). The result can be acquired in the form of a user who visited a specific facility, a user who participated in a specific event (a set of events), an event participation history of the specific user, and the like. Such an acquisition result (107) can be stored in the data storage unit (010) of the control terminal (008) as necessary, and can be used for subsequent marketing activities.

(1-2) Positional relationship between facility and location specifying terminal FIG. 2 shows a simplified locational relationship between facility and location specifying terminal. In many conventional location information collection marketing systems, a user's movement or stay at a facility is directly required from the user's location information. That is, a method of confirming the presence of the user at the corresponding location based on the user location information acquired from the location specifying terminal at the destination or the destination is adopted.

  On the other hand, in the case of the system according to the form example, the position information is indirectly performed by a position specifying terminal located at a place other than the user's destination or stay. Here, the location specifying terminal is, for example, a ticket gate at a station, an entrance / exit gate on a toll road, a product sales terminal, or the like. In the case of this embodiment, the user's movement log is recorded based on a user ID or the like collected through a location terminal as a passing point on the user's movement route. Therefore, in the case of a form example, the method of estimating a user's actual destination based on indirect information, such as event information, based on these indirect movement logs is adopted. This is the difference between the system according to the embodiment and the conventional system.

  The example of FIG. 2 will be described. Here, a case is considered where there is information that has passed through the location terminal A (451). In this case, there is a high possibility that the destination candidate to which the user has headed is either the facility 1 (461), the facility 2 (462), or the facility 3 (463) that is slightly apart. Actually, in most cases, there are a plurality of facilities near each location terminal. If the location specifying terminal and the facility have a one-to-one correspondence, and the situation is that a user who has passed a certain location specifying terminal is heading to the facility with a 100% probability, this embodiment will be described. This means that there is no need for a destination estimation function.

  However, as in the case of this embodiment, an apparatus that can acquire the user's position information only intermittently, such as a station ticket gate, a toll road entrance / exit gate, a merchandise sales terminal, etc. is assumed as the position specifying terminal. In this case, estimation processing using one or a plurality of factors such as the passage date and time of the location specifying terminal, the event date and time, the regularity of the user's behavior is required.

  In the case of FIG. 2, the user who has passed through the location terminal B (452) is likely to go to the facility 3 (463) or the facility 4 (464). In addition, the user who has passed the position specifying terminal C (453) is likely to go to the facility 4 (464) or the facility 5 (465). In addition, it is highly likely that the user who has passed the location specifying terminal D (454) headed for the facility 2 (462), the facility 5 (465), or the facility 6 (466).

(1-3) Data Structure Example Here, a data structure of typical data stored in the data storage unit (100) will be described. FIG. 3 shows the data structure of various data. The ID-attached location log data (101) includes a log ID (111), a location log target user ID (112), a location specifying terminal ID (113), and the user has passed this location (that is, processing has been performed). ) Information such as date and time (114) and its processing type (115) is included. The processing type (115) is information indicating the type of processing executed by the location specifying terminal such as “entrance” or “entrance” for a ticket gate or an entry / exit gate, and “purchase” for a settlement or product sales terminal. is there.

  User data (102) for managing information about the user includes personal information such as user ID (121), name (122), gender (123), date of birth (124), residence area (125), and membership type ( 126) and the like. Here, information that can identify an individual such as a name and a detailed address is not essential for the operation of the system. Attribute information such as gender and age is not always necessary, but it is often useful for marketing activities after data processing. Various information related to users not listed here (profession, work area, etc.) Even if it is useful for subsequent marketing activities, it is desirable to collect and store it.

  The location specifying terminal data (103) stores data relating to location specifying terminals such as ticket gates, entrance gates, and sales terminals. The position specifying terminal data (103) includes the terminal ID (131), the name of the terminal or terminal (132), such as “XX Station No. 1 ticket gate”, “△△ interchange up gate”, and the type of terminal (133). , Information such as the latitude and longitude (134) of the place where the terminal is installed, the date and time of installation (135), the expiration date or removal date and time (136), and the like. Here, the reason for storing the installation date and time (135) and the expiration date (136) is that terminals having the same ID may be moved or removed over time.

  In addition to the facility ID (141) and the facility name (142), the facility data (104) for storing data relating to the facility includes the type of facility (type of landmark such as a stadium, exhibition hall, museum, store) (143), Includes information such as latitude / longitude of the location (or its range) (144), nearest location terminal and its distance (multiple if there are multiple) (145), installation date / time (146) and expiration date (147) . Here, as shown in this figure, the information on the nearest location terminal may be stored in the facility data in advance, or the distance calculation is performed each time the extraction process described later is actually performed. You may list them. If the location terminal does not move very much and there are many absolute numbers, it is more convenient to store information about the nearest location terminal in advance in the facility data, but the location terminal moves frequently, When facility data and location terminal data are frequently updated, such as being installed, it is more convenient to list information on the nearest location terminal each time.

  Event data (105) includes event ID (151), event name (152), event type (153), venue facility ID (154), start date / time (155) and its margin (156), end date / time ( 157) and its margin (158) and various information such as information (159) on the scale of the event. Here, the margin of the start date / time and the end date / time considers not only the case where the participant arrives at the venue at the exact start time and leaves the venue at the exact end time. For example, the presence of a user who arrives early (or after a delay) in advance or returns after a little time has elapsed from the end time is considered. The margins before and after are determined according to the type and characteristics of the event. In the example of the figure, 120 minutes before and 90 minutes behind the start date and time, 30 minutes ahead and 120 minutes behind the end date and time. Therefore, if the start date is 15:00 on July 15, 2008, the margin range is from 11:00 to 14:30 on the same day. If the end date / time is 16:02 on July 15, 2008, the margin range is from 15:32 to 18:02 on the same day.

  Extraction result data (106) obtained from these data by a predetermined process includes a user ID (161), an event ID (162) considered to have participated, and the likelihood that the user has participated in the event (reliability). By collecting these data, including information such as (163), information (108) relating to a particular user's hobbies and interests, and conversely, information relating to the tendency of users who participated in a particular event (109), etc. Can be used for marketing and event planning.

(1-4) Outline of Processing Operation (a) Outline FIG. 4 shows a schematic image of a processing procedure executed in the entire system according to the embodiment. In the figure, position information collection processing (310) and event participation determination processing (320) are shown.

  In the position information collection process (310), the server process is shown in the left column, the position specifying terminal process is shown in the middle column, and the user process (or ID notification terminal process held by the user) is shown in the right column. When the user moves or acts (311), if a ticket gate or entrance / exit gate passing through a station, an interchange, a store, or a settlement process occurs (312), the user or terminal ID is notified to the location terminal Is done.

  The location specifying terminal acquires the user ID and the date and time (313), and transmits this information to the server (314). This process may be executed in real time, or may be sent periodically and collectively. The server (001) stores the received data as ID-attached position log data (101) (315). This process is performed at any time throughout the user's activity. In the figure, broken arrows indicated by bold lines indicate access from the operation terminal and display terminal to the server (001) through the network, and dotted arrows indicated by thin lines indicate database access to the server. Indicates access.

  Next, the event participation determination process (320) will be described. In the event participation determination process (320), when an operator inputs or selects a condition on the control terminal (321), an instruction is sent to the server, and the server adds log data with ID (according to a predetermined procedure described later) 101) and collation data (110) are used to collate data and extract users (322). The extraction result is stored as data (106) and transmitted to the operator. The operator confirms the result (323), inputs new conditions according to the result (321), and uses it for subsequent marketing activities. (324).

(B) Contents of collation / user extraction processing FIG. 5 shows an example of the processing procedure of the collation / user extraction processing (322) executed by the server (001) in the event participation determination processing (320). FIG. 5 shows a state in which processing (330) for extracting a user who seems to have participated in this event is executed when an event (= EventA) with an operator (003) is designated as a search condition. .

(Process 331)
First, the process (331) will be described. In the process (331), the facility ID (154) of the event venue is set to “VnID” based on the event related data (105). Also, based on the event related data (105), the time obtained by subtracting the previous margin (156) from the event start date (155) is StTime1, and the time obtained by adding the rear margin (156) to the event start date (155) is StTime2. The time obtained by subtracting the previous margin (158) from the event end date and time (157) is set to EdTime1, and the time obtained by adding the rear margin (158) to the event end date and time (157) is set to EdTime2. Also, based on the event-related data (105), a set of dates picked up at least one day at any date (usually the previous day, the next day, or the same day of another week) when no event is held is set in OtDate [] To do. The above is the processing (331).

(Process 332)
Next, referring to the nearest terminal ID (145) registered for the facility VnID (facility data (104)) that is the event venue of Event A, processing (333) and (334) described later for (= LocID) respectively. , (335) is performed (332).

(Process 333)
In the process (333), based on the distance between the venue facility set in the process (331) and the nearest terminal, the time required for the movement between the two is calculated and set as DsTime.

(Process 334)
Next, from the location log data with ID (101), the location specifying terminal ID (113) is equal to LocID, the processing type (115) is “participation” or “purchasing”, and the processing date (114) is (StTime1 -DsTime) and log data that is earlier than (StTime2-DsTime) (that is, the target location terminal is passed in the entry direction in the time zone taking into account the margin and the required time to the event venue at the event start time. Log data), and a list a ′ (181) obtained by adding a likelihood described later.

(Process 335)
Next, from the location log data with ID (101), the location specifying terminal ID (113) is equal to LocID, the processing type (115) is “entrance” or “purchase”, and the processing date and time (114) is ( Log data that is later than (EdTime1-DsTime) and before (EdTime2-DsTime) (that is, passing through the target location terminal in the direction of entry in the time zone that includes the margin and the time required from the event venue at the event end time) Log data) is extracted and the likelihood is given to this list b ′ (182).

(Process 336)
By repeatedly executing the processing (332) for each nearest terminal LocID, a list a ′ (181) and a list b ′ (182) are obtained for each nearest terminal LocID. In the process (336), those lists obtained by the same process are integrated into a list a (183) and a list b (184), respectively. Thereafter, a user ID common to the list a (183) and the list b (184) is extracted, and this is set as a list c (185).

(Process 337)
Subsequently, from the ID-added location log data (101), the user ID (112) is included in the list c, the location specifying terminal ID (113) is included in the set of the nearest terminals LocID of VnID, and the processing date and time (114 ) Extracts log data included in OtDate [] and sets it as a list d (186). This list d (186) represents a log in which a user (list c (185)) who may have participated in the event from the log on the day of the event stayed near the venue on the day without the event. . Therefore, the users included in the list d (186) are not necessarily staying near the venue for Event A, but are likely to have come near the venue for another reason.

(Process 338)
Therefore, processing for removing the list d (186) from the list c (185) is performed. In this way, a list of users (list e (187)) from which users included in the list d (186) are removed from the list c (185) is extracted as users who are highly likely to participate in Event A.

(C) Method for Improving Estimation Accuracy With the processing procedure described above, it is possible to estimate the approximate users who have participated in the event. Hereinafter, a processing method for further improving the estimation accuracy will be described. Specifically, each parameter such as the interval between the entry time and the event start time, information on which nearest terminal has been passed (distance between the venue and the location specifying terminal), and the stay time between the entry time and the entry time By defining the likelihood for event participation and assigning this likelihood to the extraction result, the accuracy of the extraction result is improved.

  FIG. 6 shows an example of weighting when the likelihood is calculated. In the processing of steps (334) and (335) in FIG. 5, from the start date / time (155) and margin (156) of the event defined by the event related data (105), the end date / time (157) and the margin (158). During the designated time periods (343) and (344), users who passed through the location terminal (006) were extracted as users who are highly likely to participate in the event.

  At the time of this extraction, likelihood is calculated by multiplying values represented by weighting curves (341), (342), and (348) as shown in FIG. In other words, in the time zone that takes into account the margin, based on the type of the event, the event-specific conditions, and the distribution of the visit time zone obtained empirically, how much before or after the actual event start time or end time Depending on whether or not (= approximately passed the location specifying terminal), the degree of possibility of participating in the event is calculated.

  In the example of FIG. 6A, the value of the weighting curve is larger at the time (345) when the user a passes the position notification terminal than at the time (346) when the user b passes. Therefore, in this case, the user a is more likely to participate in the event. For example, it is possible to improve the reliability by setting a threshold value for this likelihood and extracting only the users having the threshold value or more.

  In addition to such weighting by passing time (340), weighting by staying time (347) can be used as shown in FIG. 6B. Also in the case of FIG. 6B, the value of the weighting curve corresponding to the stay time (345) of the user a is larger than the value of the weighting curve corresponding to the stay time (346) of the user b. Therefore, the user a has a higher likelihood of participating in the event. In this case as well, reliability can be improved by setting a threshold for this likelihood and extracting only users that are equal to or greater than the threshold.

  In addition to these examples, the likelihood calculation is weighted by the nearest location terminal used (such as increasing the likelihood the shorter the distance between the location terminal and the event venue), the user is specified this time. If there is a history of participation in similar events (same event type) in addition to other events, various methods such as increasing the likelihood of participation in the designated event can be considered.

  In addition, the final event participation determination is performed by calculating the addition result of the individually weighted results for the plurality of items, and setting a corresponding threshold value in advance, so that only users who are equal to or greater than the threshold value are determined. It is also possible to apply a technique for extracting. Thus, if the determination which reflected the likelihood about several items is performed, total reliability can be improved to that extent.

(1-5) Specific Example Hereinafter, a processing procedure executed when extracting event participants from collected log data will be described based on FIGS.

  FIG. 7 is a diagram showing the movement of the user in a simplified manner. In this example, the user P (401) is a user who commute to the X office (421) near the station A every day, and the user Q (402) and the user R (403) It is assumed that the user who participated only on a certain day in the exhibition held in 422), and the user S (404) is a user who occasionally watches the game in the stadium.

  In this case, the user P (401) goes to the X office (421) every morning through the ticket gate (411) of the A station, and goes home every night through the ticket gate (411) of the A station. One day, the user Q (402) passes through the ticket gate (411) of the A station and participates in the exhibition held at the Y exhibition hall (422), but moves to the B station without returning to the A station. I passed the ticket gate (412) at station B and went home. The user R (403) passes the ticket gate (411) of the station A on the same day as the user Q, purchases drinks at the vending machine C (413) on the way, and then exhibits at the Y exhibition hall (422). I went to the office after passing the ticket gate (411) at station A in the afternoon. The user S (404) passes through the ticket gate (411) of the A station on the evening of the same day, goes to the Z stadium (423), watches the game of the night game, and then again at the ticket gate (411) of the A station on the same night. ) And went home. The user S (404) passes the ticket gate (414) of the D station to watch a game held at the W stadium (424) in another place on another day, and again after watching the game, the ticket gate (414) of the D station. ) And went home.

  Specific examples of data collected in relation to the movement of each user shown in FIG. 7 are shown in FIGS. FIG. 8 shows specific examples of user data (102), location specifying terminal data (103), and facility related data (104), which are relatively static data.

  User data (102) stores data of user P (401), user Q (402), user R (403), and user S (404). These user data are separately collected in association with the ID of the IC ticket when passing through the ticket gate of the station. However, information that can identify an individual such as a name and a residential area is not necessarily required. However, it is essential that each user (card, terminal) can be identified by an identifier such as a unique card ID or user ID.

  The location terminal data (103) stores data such as ticket gates, entry / exit gates, vending machines and merchandise terminals. These also have unique terminal IDs and are associated with position information. The facility-related data (104) is information regarding landmarks such as offices, exhibition halls, and stadiums, and holds location information of the facility itself, the ID of the nearest location specifying terminal, and the like. By using this facility-related data (104), it is possible to narrow down approximate candidates for estimating where a user's destination that has passed through a location specifying terminal near each facility is.

  For example, in the example of FIG. 8, the X office (421), the Y exhibition hall (422), and the Z stadium (423) are relatively close to the A station ticket gate (411), and in particular, from the Y exhibition hall (422) to the vending machine. It can be seen that C (413) is very close and the B station ticket gate (412) is also relatively close. It can also be seen that the nearest terminal of the W stadium (424) is the D station ticket gate.

  FIG. 9 shows specific examples of event-related data (105) and ID-attached position log data (101). The ID-attached position log data (101) is data that is added every moment, and the event-related data (105) needs to be maintained in accordance with the daily update status of event information. The event-related data (105) includes the facility ID of the venue where the event is held, in addition to the name, type, and date / time of the event. In this example, the IT exhibition where user Q (402) and user R (403) participated is the Y exhibition hall (422), which was held from morning to evening on July 15, 2008. It is shown that. It also shows that a professional baseball game was held at Z Stadium (423) on the same night. In addition, it is shown that another baseball game is being held at the W Stadium (424) on the evening of July 31, about two weeks later.

  Here, in the ID-added position log data (101), logs that the user P (401), the user Q (402), the user R (403), and the user S (404) have passed through the position specifying terminals are left. Yes. Stays near Station A can be expressed as the time from ticket gate entry to admission, and the same user entry / entry pair is the boundary of the time zone during which the user stayed in the vicinity and performed some activity. Can be considered. In the case of the data example of FIG. 9, there are three pairs of entry and entry regarding the user P (401). That is, for the user P (401), on the morning of July 15, 2008 (441), July 16, 2008 (445), and July 17, 2008 (446), the station A ticket gate (411) To enter the A Station ticket gate (411) on the same night. Similarly, it can be seen that the user Q (402) entered the A ticket gate (411) on the morning of July 15, 2008 and entered the B station ticket gate (412) on the same night (443). It can be seen that the user R (403) entered the A ticket gate (411) on the morning of July 15, 2008 and entered the A station ticket gate (411) in the early afternoon (442). There are two pairs of entry and entry for user S (414). That is, for the user S (414), the entrance / exit (444) of the A ticket gate (411) from the evening to the night on July 15, 2008, and the D ticket gate (from the evening to the night on July 31, 2008) ( 414) entry and exit (447).

  FIG. 10 shows an example of extracting users who participated in the event from the data group shown in FIGS. 8 and 9 according to the processing procedure shown in FIG. Here, among the event data shown in FIG. 9, “ΔΔ Show”, which is an IT exhibition, is referred to as Event A. In this case, first, a list a (183) and a list b (184) are acquired by the processes (331) to (335) in FIG. In the list a (183), a log that passes (enters or purchases) the nearest location identification terminal ID of the exhibition hall in the morning of the event date (time zone around the event start time considering the margin) is displayed in the list b (list b). In (184), logs are extracted that have passed (entered or purchased) the location specifying terminal ID nearest to the exhibition hall in the afternoon of the same day (time zone around the event end time considering the margin).

  Thereafter, the list c (185) is obtained as a user who is commonly included in the list a (183) and the list b (184) by the integration process of the process (336) in FIG. Here, together with the user IDs in the list c (185), the likelihood (likelihood) that the user has participated in the event is calculated. For example, as shown in FIG. 6, a weighting value based on each of the passage time of the position specifying terminal and the distance from the venue of the position specifying terminal is calculated. In the list c (185) shown in FIG. 10, the reason that the likelihood of the user R00003 is lower than that of other users is that the ticket entry time of the R00003 in the list b (184) is much before the event end time. It depends. Further, the list d (186) is obtained by the process (337) of FIG. In the list d (186), a log in which the user included in the list c (185) passes the nearest terminal ID is extracted in addition to the event date.

  Thereafter, the process (338) of FIG. 5 removes the user P00001 that is moving in the vicinity other than the event date from the three users included in the list c (185), and finally The two users shown in the list e (187) are extracted.

  As described above, in the example shown in FIGS. 7 to 10, four users are moving near the station A, but as described above, only two people, Mr. Q and Mr. R, except Mr. P who commute every day. , It will be extracted as a user who participated in the exhibition held in the exhibition hall. In addition, it is estimated that Mr. S, who has a late entry time and a late entry time at the A ticket gate, visited a game held at the stadium at night, not at an exhibition held in the daytime.

  Of course, in this example, it is assumed that the user's behavior is simple, so it was easy to determine participation in the event, but the actual movement of people is more complicated, and various factors are combined. Yes. For this reason, even if likelihood calculation is used, it is not easy to completely estimate the user's action purpose.

  Therefore, it is important to increase the accuracy by increasing the likelihood calculation conditions described above, calculating the likelihood for participation in a plurality of events, and repeating the determination process. Even in this example, the accuracy of the estimation that Mr. S watched the game is not high only with the log of July 15, 2008. However, on July 31, 2008, another log was added near another stadium, and a similar log existed on another day. However, it is possible to improve the estimation accuracy of watching the game on July 15, 2008. In this way, by taking into account the history of a certain user regularly visiting a specific place and the history of repeatedly participating in similar events, the user's behavior tendency is clarified The estimation accuracy can be increased. As a result, information useful for marketing activities can be extracted.

(1-6) Screen Example Hereinafter, based on FIGS. 11 to 15, a display screen example of a control terminal used in the information processing system according to the embodiment will be described. These screens are examples of screens displayed by the operator (003) in FIG. 1 through operation inputs to the control terminal (008). For example, it is an example of a screen that is displayed when a search condition is input or selected by an operator (003), and is an example of a screen that is used to display an extraction result for the search condition.

  FIG. 11 is an example of a screen used when selecting an event. The event selection screen (510) displays event candidates (514) extracted according to preset conditions. When an event is selected, the event confirmation button (515) is pressed to determine the event. When the event search condition input button (511) is pressed, a screen (520) for separately setting conditions is displayed. It is also possible to input an arbitrary keyword (512) and search using this keyword (513). When selecting a plurality of events or setting more detailed conditions, a plurality of selection / detailed condition buttons (517) can be pressed to move to a detailed condition setting screen (not shown). If it is desired to cancel the selection of the event, a cancel button (516) is pressed.

  On the event search condition input screen (520), various conditions such as a rough genre (521), a more detailed genre (522), a holding period (523), and a holding area (524) are selected from a list box or the like, and an event search button Press (525) to display the target event candidates. If it is desired to cancel the input, a cancel button (526) is pressed.

  FIG. 12 is an event participation user extraction result screen (530) used when displaying the extraction result of the user who participated in the designated input event. When an event is selected through the work screen shown in FIG. 11, the server (001) extracts a user who seems to have participated in this event by the extraction process described above, and the possibility that the user has participated in the event (likelihood). ). In the event participation user extraction result screen (530), information (531) registered for the event is displayed, and an event details screen (570) shown in FIG. 14 is displayed by pressing an event details information confirmation button (532). You can check the detailed information of the event.

  In a list (533) of user lists that are considered to have participated in the event, user ID, likelihood, and other summary information about other users are displayed. It is also possible to execute sorting for each item. A button (534) for confirming detailed information of each user is arranged in the user display column. When this button is pressed, the screen shifts to a user information detail screen (550) shown in FIG. The event participation user extraction result screen (530) also displays a narrowing condition setting button (535) for setting user narrowing conditions to be displayed in the list (533). When this button is pressed, a narrowing condition can be set on the next narrowing condition setting screen (540). When returning from the event participation user extraction result screen (530) to the event selection screen (510), the return button (536) is pressed.

  The narrow-down condition setting screen (540) includes a minimum value (541) and maximum value (542) of likelihood of event participation, a minimum value (543) and maximum value (544) of age, designation of sex (545), (546). ), (547), residential area (548), (549), etc. are specified, and the narrowing-down conditions for the user to be displayed are set. When the condition confirmation button (537) is pressed with the setting completed, the user list (533) in the event participation user extraction result screen (530) is updated. Note that if the user wants to cancel the setting condition on the narrowing-down condition setting screen (540), a cancel button 538 is pressed.

  FIG. 13 is an example of a screen focused on a specific user. In the event participation user extraction result display screen (530) of FIG. 12, an example of a screen displayed when a button (534) for confirming details of each user is pressed is a user information detail screen (550). . Registered information such as the user ID (551), name or pseudonym (552), sex (553), age (554), residence (555), member type (556) of the selected user is displayed. Here, when the button (557) "View other events that this person participated" is pressed, the participation event estimated through the extraction process of the server (001) is extracted, and the user may have participated in the event. Displayed with (likelihood).

  On the participation event extraction result display screen (560) designating the user, information (561) about the user is displayed, and the user detailed information can be confirmed by pressing the user detailed information confirmation button (562). In the list (563) of the event list that the user seems to have participated, the event name, likelihood, and other summary information about the event are displayed. It should be noted that display items can be rearranged and displayed by executing sorting for each item. A button (564) for confirming detailed information is arranged in each event column. When this button is pressed, the event detailed information screen (570) shown in FIG. 14 is displayed. Similarly to the narrowing down of the user, the narrowed down condition setting button (565) can be pressed to change to the narrowed down condition screen (580) shown in FIG. 14, and the extracted events can be narrowed down.

  FIG. 14 is an example of a screen focusing on a specific event. On the event details screen (570), detailed information such as the event ID (571), event name (572), event type (573), date and time (574), event venue (575), etc. of the selected event can be confirmed. Here, when a button (576) "View users who participated in this event" is pressed, a list of users who participated in the event can be referred to by the event participation user extraction result screen (530) shown in FIG. it can. When the return button (578) is pressed, the display can be returned to the participation event extraction result display screen (560).

  In the narrow-down condition setting screen (580), the likelihood (581), (582), type (583), event period (584), (584), (from the participation event extraction result screen (560) displayed for a certain user shown in FIG. 585), the event area (586), (587), etc., conditions for narrowing down the events are set. When each condition is specified and the “Confirm condition” button (588) is pressed, the narrowed-down events are displayed on the participation event extraction screen (560). Also in this screen, when the cancel button (589) is pressed, the state immediately before the participation event extraction result screen (560) is returned.

FIG. 15 shows the transition relationship between the screens shown in FIGS. 11 to 14 described above. As shown in FIG. 15, it can be seen that when the button on each screen is pressed, the display screen shifts to another screen and the result can be referred to while changing the conditions.
(1-7) Effects of the form example As described above, when the system according to the form example is used, the history information regarding the movement intermittently acquired from the passing point in the process of the user's movement is selected as the population. It is possible to extract a user who is estimated to have participated in another event (or a set of events) with high accuracy. That is, the user's behavioral purpose and the user's behavior tendency by paying attention to the selected event information etc. while making the log data of the IC ticket including the purchase information at the vending machine and the time of passing the ticket gate of the station as a population. Can be analyzed indirectly.

  In addition, since users of various purposes pass through the entrance / exit gate, the likelihood of event participation is generally low only by one estimation process. However, by repeating this estimation process, the user enters and exits. The estimated likelihood of the original target that has passed through the gate can be gradually improved.

  In addition, the extracted users can be displayed by being rearranged according to various conditions, or can be displayed after being narrowed down. Thereby, the tendency of the user layer who participated in the event can be found. On the contrary, a specific user (or a set of users) can be selected to extract events in which the user has participated, rearranged according to various conditions, or narrowed down and displayed. Thereby, the hobby and interest hidden in each user can be estimated based on the event participation tendency of the selected user.

  These results can be used in a wide range of fields such as marketing activities, event planning and publicity activities.

  In the above description of the embodiment, a single event and a single user have been extracted for the sake of simplicity, but of course, similar processing is performed for targets such as a set of multiple events and a set of users. Can be executed. In addition, by executing the processing, it becomes possible to grasp the participation tendency of the user and the set of events with higher accuracy.

  In addition, the information processing system described in this embodiment has an advantage that psychological resistance to the collection of the user's position information can be reduced because the user's position information is not always acquired. In addition, the collection of location information associated with the passage of ticket gates, entrance gates, fee gates, etc., purchases, etc. does not require the user to perform additional actions beyond general actions. Therefore, it is possible to collect position information for many users. In addition, in the case of this method, it is not necessary to prepare a special terminal for acquiring position information, and an effect that the installation cost of the position specifying terminal can be suppressed can be expected.

  In the case of the first form example, the example in which the facility where the input event is held is automatically specified has been described. However, the facility (that is, the place) can be directly input. In this case, it is also possible to execute an analysis process using various data associated with the facility (for example, event, available time, etc.) as additional narrowing conditions.

(2) Example 2
Next, a second embodiment of the information processing system will be described based on FIG. The basic system configuration and processing operation are the same as those in the first embodiment.

  In this embodiment, a case will be described in which attention is paid to purchasing behavior associated with the movement of the user. FIG. 16 shows a simplified behavior example of the user assumed in this embodiment. FIG. 16 corresponds to the behavior example of FIG.

  The difference between FIG. 16 and FIG. 7 is that the location specifying terminal is a product sales terminal. In the case of the first embodiment, the user's position information is collected mainly by the transportation IC ticket and the ticket gate. However, in this embodiment, a terminal or a point card storing electronic money is used. Assume that the provision of the used user ID and the settlement process are executed simultaneously.

  In the case of this embodiment, when a payment terminal (or a card processing terminal may be separately prepared) at each store receives electronic money or a point card held by the user, the user ID is recorded along with the payment contents. Since each payment terminal can refer to the location information of the installed store, it can be regarded as the location specifying terminal in FIG.

  FIG. 16 shows the movement history of three people. One person is Mr. P (477) who works in the office (474). Mr. P (477) visits convenience store A (471) during the lunch break and purchases lunch, then returns to the office and buys a drink at vending machine B (472) on his way home. Another person is Mr. Q (478). Mr. Q (478), after purchasing drinks at vending machine B (472), participated in an exhibition held at Y Exhibition Hall (476), and then returned to Cafe C (473) to drink coffee. I'm back. The other is Mr. R (479). Mr. R (479), after killing time at Caffe C (473), watched the game at Z Stadium (475) and returned with a warm drink at vending machine B (472). At this time, a purchase log (470) as shown in the lower part of FIG. 16 is collected and stored in the data storage unit 100.

  In the case of general customer information analysis, the purchase log (470) itself is used as a record of the purchase behavior of the user. For example, Mr. Q's total daily purchase amount is 500 yen, Mr. R's purchasing activities are concentrated after the evening, Mr. P uses convenience store A several times a week, In B, it is used for data analysis such as selling a warm drink after the evening, and such analysis is used for marketing.

  However, in the case of this example, we take one step further from such analysis and focus on what was the real purpose of Mr. Q coming to this place. In other words, by collating with separately prepared event information, facility information, etc., the user's behavioral form buried in the intermittently obtained purchase data is also analyzed, and the result is used for marketing activities For the purpose.

  Since the content of the specific purpose estimation process is the same as in the first embodiment, it will be omitted, but it is possible to estimate each user's destination with high accuracy, so that Mr. P is a person commuting to this area. , Q is a person who visits IT-related exhibitions, so he will participate in the next similar event, or Mr. R often sees a team's matches, so he is interested in that team and sports. It is possible to estimate the interests and hobbies hidden in the moving behavior, such as a certain person, and to highlight them.

Based on the position information of the user obtained intermittently in a form close to unconsciousness, it is possible to estimate the user's movement, stay, and participation in various events. This makes it possible to estimate not only the user's destination but also the user's unique interests and interests hidden by the movement and behavior accumulated intermittently. Participating in hobby events and staying in facilities that are particularly difficult to create a list of participants is useful for planning the next event and publicity activities by knowing the attributes and movement trends of the participants. Can do. In addition, marketing activities can be performed according to the user's hobbies and interests.
(3) Other Embodiments In the case of the above-described embodiment examples, a case where the user's action purpose, hobbies, etc. are indirectly estimated based on the user's action collected intermittently or history information accompanying movement will be described. did. At this time, the case where the user or history corresponding to the search condition is displayed as a list has been described, but only the total number corresponding to the search condition may be displayed.

  001: Server, 002: User, 003: Operator, 004: Network, 005: ID notification terminal (card), 006: Location specifying terminal, 007: Facility, 008: Control terminal, 009: Data storage unit of location specifying terminal 010: Data storage unit of control terminal, 100: Data storage unit, 101: Location log data with ID, 102: User data, 103: Location specifying terminal data, 104: Facility related data, 105: Event related data, 106 : Extraction result data, 107: acquisition result, 108: hobby / interest of a user, 109: event participant tendency, 110: data for verification, 111: log ID, 112: user ID, 113: location terminal ID, 114: processing date and time, 115: processing type, 121: user ID, 122: name, 123: gender, 124 Date of birth, 125: Residence area, 126: Member type, 131: Terminal ID, 132: Location / terminal name, 133: Type, 134: Latitude and longitude, 135: Installation date and time, 136: Expiration date (removal date), 141 : Facility ID, 142: Facility name, 143: Type, 144: Latitude and longitude, 145: Nearest terminal ID and distance, 146: Installation date and time, 147: Expiration date (removal date), 151: Event ID, 152: Event name, 153: Type, 154: Facility ID of the venue, 155: Start date / time, 156: Start date / time margin, 157: End date / time, 158: End date / time margin, 159: Participation scale, 161: User ID, 162: Event ID, 163: Likelihood, 181: List a ′, 182: List b ′, 183: List a, 184: List b, 185: List c, 186: List d, 187: List e, 200: Control unit, 201: Location log-event collation processing unit, 202: Data maintenance processing unit for collation, 203: Data communication / input / output processing unit, 204: Database access processing unit, 301: Location information transmission processing, 302: Condition input (selection) processing, 303: Result acquisition processing, 304: Event participation, 310: Location information collection processing, 311 to 319: Processing steps, 320: Event participation determination processing, 321 to 324, 330 to 338: processing step, 340: example of weighting by passage time, 341 to 342: weighting curve, 343 to 344: time zone taking into account margin, 345: staying time near event venue of user a, 346: user b Stay time near the event venue, 347: example of weighting by stay time, 348: weight Marking curve, 401: User P, 402: User Q, 403: User R, 403: User S, 411: A station ticket gate, 412: B station ticket gate, 413: Vending machine C, 414: D station ticket gate, 421: X office , 422: Y exhibition hall, 423: Z stadium, 424: W stadium, 441-447: entry / entry pairs, 451-454: location terminal, 461-466: facility, 470: purchase log data, 471: convenience store A, 472: Vending machine B, 473: Cafe C, 474: X office, 475: Z stadium, 476: Y exhibition hall, 477: User P, 478: User Q, 479: User R, 500: Terminal screen for control, 510: Event selection screen, 511: Search condition input button, 512: Keyword input box, 513: Keyword search button, 514: Event Complementary list, 515: Event confirmation button, 516: Cancel button, 517: Multiple selection / detailed condition setting button, 520: Event search condition input screen, 521: Large genre selection list box, 522: Small genre selection list box, 523: Event period selection list box, 524: Event area selection list box, 525: Event search execution button, 526: Cancel button, 530: Event participation user extraction result screen, 531: Event overview, 532: Event details confirmation button, 533: Event Participating user list, 534: user details confirmation button, 535: narrowing condition setting button, 536: cancel button, 537: narrowing condition confirmation button, 538: cancel button, 540: narrowing condition setting screen, 541: likelihood minimum value setting button 542: Maximum likelihood setting box, 543: Minimum age setting box, 544: Maximum age setting box, 545: Gender condition (all), 546: Gender condition (male only), 547: Gender condition (female Only), 548 to 549: residence area setting box, 550: user information detail screen, 551: user ID, 552: user name, 553: gender, 554: age, 555: residence, 556: member type, 557: participation Event search button, 558: Cancel button, 560: User-specific participation event extraction result screen, 561: User overview, 562: User details confirmation button, 563: Participation event list, 564: Event details confirmation button, 565: Refinement condition setting button , 566: Cancel button, 570: Event details screen, 571: Event ID, 5 72: Event name, 573: Event type, 574: Event date, 575: Venue, 576: Event participation user search button, 577: Cancel button, 580: Refinement condition setting screen, 581: Minimum likelihood setting box, 582: Likelihood maximum value setting box, 583: Event type setting box, 584 to 585: Holding period setting box, 586 to 589: Holding region setting box, 588: Condition confirmation button, 589: Cancel button.

Claims (8)

A storage device for storing history information composed of identification information of a moving body whose position information acquisition apparatus has been confirmed, confirmed position information, and confirmed date and time information;
Only the history information for each moving body collected by the position information acquisition device located around the place input by the operator as the search condition or the place corresponding to the event input by the operator as the search condition is extracted from the storage device. By comparing the processing and the extracted history information for each moving body with at least one of the stay time, the past behavior history, the distance from the place input by the operator, or the time when the event was held, An information processing system comprising: a verification processing device that executes a second process for estimating a mobile object that is highly likely to satisfy a search condition. The information processing system according to claim 1, wherein in the second process, an evaluation process is executed based on a likelihood that gives a certainty of a collation result. The information processing system according to claim 2, wherein the verification processing device displays an estimation result corresponding to the search condition as a list or a total number of moving objects. The information processing system according to claim 3, wherein when a specific moving object is specified in the list, a past action history related to the moving object is displayed. A process of accepting a place or event as a search condition through an operation input to an input device by an operator;
Location information entered by the operator as a search condition in the case where history information composed of identification information of a mobile body whose position information acquisition apparatus has been confirmed, confirmed position information, and confirmed date and time information is stored in a storage device or A process of extracting only the history information for each moving body collected by the position information acquisition device located around the place corresponding to the event input by the operator as a search condition from the storage device;
Corresponding to the search condition by collating the extracted history information for each moving body with at least one of the stay time, past action history, distance from the place entered by the operator, or the time when the event was held And a process of estimating a moving object that is likely to be performed. The information processing method according to claim 5, wherein in the collation process, an evaluation process is executed based on a likelihood that gives a certainty of a collation result. The information processing method according to claim 6, wherein the estimation result as a result of the collation processing is displayed as a list or total number of moving objects as an estimation result corresponding to a search condition. The information processing method according to claim 7, wherein when a specific moving object is specified in the list, a past action history related to the moving object is displayed.

Images