JP5519431B2 - Mobile object feature description device, mobile object feature description method, mobile object feature description program, and recording medium - Google Patents

Description

  The present invention relates to a technique for describing characteristics of a mobile body used for selecting content to be distributed in a content distribution service for digital signage on a mobile body such as a train.

  Digital signage is a content display device such as an advertisement installed outdoors (in a moving body such as a train, in a street, in a store, etc.). For example, in a station concourse or platform, operation information, announcements, advertisements, etc. are displayed on a liquid crystal panel, or advertisements, etc. are displayed on a small liquid crystal panel in a display shelf of a retail store. . Digital signage is mainly composed of a terminal having a video display function such as a display and a personal computer, and these may be managed by a plurality of servers. In recent years, the number of digital signage installations has increased, and the number of cases in which digital signage is connected to a network for remote management and operation is increasing.

  On the other hand, in mobile terminals such as mobile phones, broadband usage such as 3G and HSDPA has become common, and it has great potential as an outdoor advertising medium, such as displaying information on the stay area on the mobile terminal. .

  It is said that it is easy to lead the recency effect as a feature of outdoor advertising media. The recency effect means that a consumer can expect an effect of promoting the purchase of a product by contacting an advertisement near a purchase point. For example, it is possible to effectively appeal to viewers during shopping by displaying product advertisements on display shelves in the store.

  When distributing content such as advertisements to digital signage and mobile terminals, if digital signage and mobile terminals are regarded as “displays” that display advertisements, etc., the characteristics of people around the display such as signage installation locations and mobile terminal users It is important to understand. As the number of displays increases, it is desirable that features are automatically understood. This is because advertisements and the like are often created and presented on the assumption of a person who views them.

  As a method for grasping the characteristics of a person around the display, a method for estimating the characteristics of a person who goes around a certain display installation location from surrounding environment information such as the surrounding facility and land price has been proposed (Non-Patent Document 1). .

Maehashi Yoshibayashi, 3 others, “Automatic generation of“ location metadata ”using estimation of visitor characteristics”, IEICE Technical Report, IEICE, March 2010, Vol. 109, No. 450, p. 49-54 Yuichiro Miyazaki and three others, “Development of behavior estimation technology for service realization according to user behavior”, NTT Docomo Technical Journal, October 2009, Vol. 17, No. 3, p. 55-61

  However, the conventional method of grasping the characteristics of a person has a problem that only a display installed in a fixed place can be handled, and a display installed in a moving body such as a train cannot be handled.

  In addition, when placing an advertisement for individuals on a mobile terminal, there is a case where it is desired to use a tendency of a plurality of users for which data has already been obtained in an initial stage where data about the user is insufficient. However, since the mobile terminal moves with the user, there is a problem that the conventional method cannot be used.

  By using the technique of Non-Patent Document 2, it is possible to estimate the behavior / situation at the end point of the individual's movement from the position history. It can be considered as a feature around the display. However, in the case of dynamic information such as estimated behavior / situation instead of registered profile information such as gender and age, simply summing up personal information is easily affected by estimation errors, so it is possible to grasp the tendency of the group. There is a problem that it is difficult.

  The present invention has been made in view of the above, and an object of the present invention is to obtain characteristics of a user who uses a mobile object that can be used for content distribution such as advertisements to the mobile object.

A moving body feature description device according to a first aspect of the present invention uses a position history acquisition unit that reads a history of position information from a position history storage unit that stores a history of position information of a user, and a history of the position information. The travel route estimation means for estimating the travel route of the user, the travel method determination means for determining the travel method of the user using the history of the location information, and the travel method uses transportation. A movement section estimation means for estimating a movement section in which the user uses the transportation means using the movement route, and an action situation of the user at the end point of the movement section using the history of the position information estimating and a behavior state estimation means for storing the movement history storage means in association with the behavioral state and the movement section, from the travel history accumulation unit, subject describe the features of a mobile of the transportation agencies An action status acquisition unit that reads out the action status associated with the moving section that falls in between, a tabulation unit that counts up the behavior status read out by the behavior status acquisition unit, and a count result of the tabulation unit as the target section And description means for storing in the moving object feature accumulating means as the characteristics of the moving object.

A moving body feature description method according to a second aspect of the present invention is a moving body feature description method executed by a computer, and reads out the position information history from a position history storage means for storing a user's position information history. A step of estimating a travel route of the user using the history of the location information, a step of determining a travel method of the user using the history of the location information, and If used, the step of estimating the movement section where the user uses the transportation means using the movement route, and the end point of the movement section of the user using the history of the position information estimating a behavioral state, a step of storing the movement history storage means in association with the behavioral state and the movement section, from the travel history accumulation unit, the mobile of the transportation agency JP Wherein the step of reading the behavioral state associated with the movement section corresponding to the write target section and a step of aggregating the action status read in step of reading the behavioral state, the counting result of said step of aggregate And a step of accumulating in the moving object characteristic accumulating means as the characteristic of the moving object in the target section.

  A moving object feature description program according to a third aspect of the present invention causes a computer to execute the moving object feature description method.

  A recording medium according to a fourth aspect of the present invention is characterized in that the above-mentioned moving body feature description program is recorded.

  ADVANTAGE OF THE INVENTION According to this invention, the characteristic of the user using a mobile body which can be utilized for content delivery, such as an advertisement to a mobile body, can be acquired.

It is a functional block diagram which shows the structure of the mobile body characteristic description apparatus in this Embodiment. It is a figure which shows the example of the log | history of a positional information. It is a figure which shows the example of a personal action information log | history. It is a figure which shows the example of the location feature in which the function division was described. It is a figure which shows the example of a mobile body characteristic. It is a flowchart which shows the flow of the process which estimates a movement area. It is a flowchart which shows the flow of the process which estimates a user's action and condition. It is a flowchart which shows the flow of the process which describes the characteristic for every area.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a functional block diagram showing a configuration of the moving object feature description device according to the present embodiment. The moving body feature description apparatus shown in the figure assumes a public transportation such as a train or a bus as a moving body, and describes a user's characteristics for each moving section and time zone. The moving section is not necessarily the starting station to the ending station of the operating section, but refers to a part of the operating section such as “Shibuya-Nakameguro”.

  1 includes a position history storage unit 301, a personal position history acquisition unit 302, a movement mode determination unit 303, a movement route estimation unit 304, a movement section estimation unit 305, a personal behavior situation estimation unit 306, Determination unit 307, operation section position information storage unit 308, personal behavior status history storage unit 309, transition probability calculation unit 310, section selection unit 101, personal behavior status correction unit 102, totaling unit 103, location feature use unit 104, location feature The storage unit 105, the moving object feature description unit 201, and the moving object feature storage unit 202 are provided. Each unit included in the moving body feature description device may be configured by a computer including an arithmetic processing device, a storage device, and the like, and the processing of each unit may be executed by a program. This program is stored in a storage device included in the mobile feature description apparatus, and can be recorded on a recording medium such as a magnetic disk, an optical disk, or a semiconductor memory, or provided through a network.

  The mobile object feature description apparatus according to the present exemplary embodiment includes a personal action situation calculation unit 300 that calculates a personal action situation history from a personal location information history and a personal action situation history to calculate a mobile object feature. The calculation unit 100 is divided. Hereinafter, each part will be described.

  The position history accumulation unit 301 accumulates a plurality of individual position information histories. The location history storage unit 301 is connected to a network, acquires location information from a mobile terminal having a location detection function such as a Global Positioning System (GPS), and stores the location information.

  The personal location history acquisition unit 302 acquires a location history for each individual from the location history storage unit 301.

  The movement mode determination unit 303 determines the movement mode (walking, train, car, etc.) of the corresponding individual using the position history acquired by the individual position history acquisition unit 302.

  The movement route estimation unit 304 estimates the movement route of the individual using the position history acquired by the individual position history acquisition unit 302.

  The travel section estimation unit 305 includes a travel section position information storage unit 308 that stores the travel mode determined by the travel mode determination unit 303, the travel route estimated by the travel route estimation unit 304, and the travel route position information of public transportation. Refer to and estimate the transportation name (moving body name) and moving section when the individual uses public transportation.

  In the movement section estimated by the movement section estimation section 305, the personal action situation estimation unit 306 estimates the action / situation of the individual at the start and end points of the movement section, and together with the user ID for identifying the individual, the movement section, The estimated action situation at the moving time zone, the start point and the end point is accumulated in the individual action situation history accumulation unit 309 as the individual action situation history.

  The accompanying determination unit 307 examines an individual who has moved in the same movement section as the movement section estimated by the movement section estimation unit 305, determines the presence / absence of the accompanying person, and determines the presence / absence of the accompanying person in the personal action situation history storage unit 309. Add to the personal behavior status history.

  The operation section position information storage unit 308 stores the operation section of the moving body (train, bus, etc.) and the position information thereof. For example, a section may be defined by dividing an operation section for each station having a large number of passengers on a certain train.

  The personal behavior status history storage unit 309 accumulates a personal behavior status history in which a movement section, a movement time zone, an estimated behavior status at a start point and an end point, and the presence or absence of a companion are described for each series of movements of a person using a mobile object. .

  The transition probability calculation unit 310 calculates the transition probability of the estimated behavior status at the end point for each estimated behavior status at the start point for the estimated behavior status at the start point and the end point accumulated in the personal behavior status history accumulation unit 309.

  The section selection unit 101 selects a target section and a time zone in which features are described, and acquires a personal behavior status history using the corresponding section in the corresponding time zone from the personal behavior status history storage unit 309.

  The personal behavior status correction unit 102 corrects the estimated behavior status of the corresponding user based on the estimated behavior status of the accompanying person for a certain individual in the personal behavior status history acquired by the section selection unit 101.

  The counting unit 103 acquires the transition probability of the estimated behavior status from the transition probability calculation unit 310, and underestimates the behavior status with a low transition probability in the corresponding section as an outlier.

  The location feature use unit 104 refers to the location feature accumulation unit 105 and corrects the action situation totalized by the totalization unit 103 in accordance with the functional classification of the section end point.

  The place feature accumulation unit 105 accumulates the function classification of the place.

  The moving body feature description unit 201 describes the feature of the section of the corresponding moving body in the moving body feature storage unit 202 in a determined format according to the behavior situation corrected by the location feature use unit 104.

  The moving body feature accumulation unit 202 accumulates the features of the moving body for each section.

  FIG. 2 is a diagram illustrating an example of data stored in the position history storage unit 301. As illustrated in FIG. 2, the position history accumulation unit 301 accumulates position information including a position and an observation time transmitted from a portable terminal carried by each individual in association with a user ID. In addition, a measurement error index indicating the reliability of data may be added. In addition, when there are a plurality of types of mobile terminals capable of transmitting position information, a similar database may exist for each device.

  FIG. 3 is a diagram illustrating an example of data stored in the personal action status history storage unit 309. As shown in FIG. 3, the personal behavior status history accumulated by the personal behavior status history accumulating unit 309 includes a movement section, a day of the week division, a time zone, a behavior status, accompanying information, and a user ID. The accompanying person ID in the figure is the user ID of the accompanying person who is estimated to accompany the user in the corresponding column, and the accompanying probability indicates the probability that the estimated accompanying person ID will accompany.

  FIG. 4 is a diagram illustrating an example of data stored in the location feature storage unit 105. The location feature accumulation unit 105 accumulates the function classification for each area. Each area may be defined by a center point and a distance such as a radius of 500 m from a center point such as a station at the end point of the moving section. As for functional classification, for example, it is classified into two types, a multi-functional area and a single-functional area. The proportion of land use classification within the defined area is examined. If the proportion of commercial land is high, it may be a multi-function area, otherwise it may be a single function area.

  FIG. 5 is a diagram illustrating an example of data stored in the moving object feature storage unit 202. In FIG. 5, for each moving object, the start point / end point of each of the N sections and the movement purpose (feature) for each time zone are stored.

  Next, the operation of the moving object feature description device in this embodiment will be described.

  FIG. 6 is a flowchart showing a flow of processing for estimating the movement section of each individual.

  First, the personal position history acquisition unit 302 refers to the position information stored in the position history storage unit 301 and acquires the position information of the individual whose movement section is to be estimated based on the user ID associated with the position information ( Step S101). The position information stored in the position history storage unit 301 is collected by a method such as periodically uploading GPS logs using, for example, a GPS function and dedicated software installed in a mobile phone. Such data is managed in association with a terminal ID or the like (Japanese Patent Laid-Open No. 2001-306604).

  Subsequently, the movement route estimation unit 304 estimates the movement route based on the position information acquired by the personal position history acquisition unit 302 (step S102). If the position information including latitude / longitude information is highly accurate, the movement route can be easily calculated. However, since there may be measurement errors in the location information, the history of individual location information is clustered, and the location where a plurality of location information stays in a certain area within a certain time is extracted as a “staying place”. By using a sequence as a travel route, it is possible to extract the travel route by reducing the influence of measurement errors (Nishino et al., “Examination of behavior pattern extraction method from stay location transition information”, IPSJ Research Report, 2008-UBI -20 (10)).

  Further, the movement mode determination unit 303 determines the movement mode of the target individual (whether it is a transportation such as walking or train) based on the position information acquired by the personal position history acquisition unit 302 (step S103). This includes calculating the average speed, average acceleration, maximum speed, and maximum acceleration within a certain time from the history of position information, and by combining these calculated values, such as stay / walk, movement by car, movement by train, etc. Determine the mode (Aoki et al., “Examination of moving mode determination from location log data for life log”, IPSJ Research Report, 2008-DD-67 (2)).

  Subsequently, when the movement mode determination unit 303 determines that the movement mode is movement using public transportation such as a train, the movement section estimation unit 305 is based on the movement route estimated by the movement route estimation unit 304. A movement section is estimated (step S104). For example, when the travel mode determination unit 303 determines that the travel mode from 2010/3/13/15: 00 to 2010/3/13/15: 25 is “train”, the travel section estimation unit 305 The moving body name and the operating section that best match the moving path are obtained from the moving path and the operating section position information stored in the operating section position information storage unit 308. The travel route estimated by the travel route estimation unit 304 is collated with a route map or the like, and it is determined that the corresponding operation section of the nearest route is used among routes located at a distance equal to or less than the threshold. As a result, the moving body name and the moving section are obtained, such as “Tokyo Shiranami Line” and “Shibuya Station to Shiranami Station”. The obtained moving body name and moving section are held in the storage means held by the moving section estimation unit 305 together with time information, user ID, and the like. For this purpose, a method of correcting a movement route on the nearest road using a road map using a movement route calculated from a position history on the road using an individual's walking movement (Nakamura et al., “ Similar to “Study on Automatic Extraction of Visited Stores by Analysis of GPS Log”, Proceedings of the 72nd Information Processing Society of Japan National Conference, Volume 3), the movement route obtained in step S102 is compared with the route map, etc. Among routes located at the following distances, the travel route is corrected on the nearest route, and it is determined that the corresponding operation section of the relevant route is used.

  The above processing is performed for each individual movement of each individual, and the moving body name, moving section, and time information are acquired. Thereby, the process using an individual's movement history is attained for every operation section and each time slot | zone of a moving body.

  FIG. 7 is a flowchart showing a flow of processing for estimating a user's behaviour / situation. After estimating the movement section of each individual, the behavior / situation of each individual at the start point / end point of the movement section is estimated.

  First, the individual behavior situation estimation unit 306 analyzes the movement history of each individual and estimates the behavior / situation at the start point and end point of the movement section (step S201). For this, use the stay location, time, weekday / holiday distinction, stay frequency, stay time, for example, if you stay regularly for a long time on weekday nights and holidays, it is estimated that you are “at home” . Similarly, the behavior / situation can be estimated as “going to work”, “going out” or the like (see Non-Patent Document 2).

  Then, the estimated behavior status estimated by the personal behavior status estimation unit 306 is stored in the personal behavior status history accumulation unit 309 together with the moving body name, the movement section, the time information, and the user ID (step S202). For example, as “Tokyo Shiranami Line” and “Shibuya Station to Shiranami Station”, the behavior / situation of the estimated behavior status at the start point is “working”, the estimated behavior status at the end point is “working”, etc. . The behaviour / situation to be defined depends on the method to be used and the feature to be finally adopted as the moving body feature. For example, actions / situations such as “shopping”, “meal”, and “leisure” may be used.

  Similarly, another user's movement history is processed and stored in the personal action status history storage unit 309.

  The accompanying determination unit 307 analyzes each person's movement history and checks whether there is an accompanying person in the movement section (step S203). For example, in the corresponding section, the similarity between the target individual and the position history owned by the other user who has passed the corresponding section is calculated using the positional information (for example, latitude and longitude transmitted by a GPS satellite) and the measurement accuracy of the positional information ( For example, it is calculated using an index of measurement accuracy transmitted by a GPS satellite), and if the similarity of the position history is high, it is determined that the accompanying probability is high, and the accompanying probability and another user's ID are added to the personal action situation history. (Seko et al., “Accompanied Judgment Method Considering Error Information”, IPSJ Research Report, 2008-UBI-20 (11)).

  By the above process, the process using the individual action / situation can be performed for each operation section and each time zone of the moving body.

  FIG. 8 is a flowchart showing a flow of processing for describing features for each section using the personal action situation history. The mobile body feature calculation unit 100 performs the following processing using the personal behavior status history for one week stored in the personal behavior status history storage unit 309 once a day, and updates the mobile feature once a day. Of course, the update cycle can be arbitrarily set.

  First, the section selection unit 101 selects a section, day of the week, and time zone of a moving body for which a moving body feature is to be described, and acquires estimated behavior status and accompanying information (step S301). For example, if the moving body is “Tokyo Shiranami Line”, the section is “Shibuya to Shiranami”, the day of the week is “Weekday”, and the time zone is “Morning”, the estimated behavior status and accompanying probability and accompanying ID are personal actions Obtained from the status history storage unit 309.

  Subsequently, the personal action situation correction unit 102 adjusts the weight when the estimated action situation is totaled using the accompanying probability acquired by the section selection unit 101 and the accompanying person ID (step S302). Specifically, when the accompanying probability is high (e.g., 0.80 or more), the accompanying action ID is referred to, and the estimated action status of the accompanying person for the corresponding period is acquired. If accompanied, the behavior / situation of all the accompanying persons is considered to be the same. Therefore, if the estimated behavior situation differs among the accompanying persons, the estimated behavior situation is corrected or underestimated. If more than half of the companions have the same estimated behavior status, the estimated behavior status of all the companions is corrected to that value. Otherwise, it is determined that the reliability of the estimation result is low, and the weight at the time of counting in the counting unit 103 is reduced. For example, the estimation result of each individual is underestimated as (1 / the number of accompanying persons). Thereby, the influence of an estimation result with low reliability can be reduced at the time of aggregation.

  Subsequently, the totaling unit 103 tabulates the estimated behavior status by excluding those with a low transition probability from the estimated behavior status acquired by the section selection unit 101 (step S303). For example, the totaling unit 103 refers to the transition probability calculating unit 310 and excludes the transition probability having a predetermined threshold value of 0.05 or less and does not use it for the totaling. Alternatively, based on the transition probability of the transition probability calculation unit 310, the deviation from the transition probability reference for each section is calculated, and those having a large deviation are treated as features of the corresponding section, and are low in the transition probability calculation unit 310. Even transitions with transition probabilities are not excluded. As described above, after performing the correction to exclude the estimated action situation that is highly influenced by the outlier and the estimation error, the estimated action situation in the corresponding section, day of the week, and time zone is “working” is 13 people, “shopping” "2", and so on.

  Note that the transition probability calculation unit 310 refers to the personal behavior status history accumulation unit 309, acquires, for example, the estimated behavior status of the start and end points in all sections of one week, and calculates the transition probability of the estimated behavior status. For example, out of the estimated values where the start point is “working”, the transition probability is calculated for each action / situation at the start point, such that the end point “working” is 0.16 and the end point “shopping” is 0.32. Hold it.

  Then, the place feature use unit 104 refers to the function category of the place based on the estimated action status tabulated by the tabulation unit 103, and determines the feature of the section (step S304). Specifically, if the function category of the end point area is “multifunctional”, the proportion of the estimated action status is calculated. If the function category is “single function”, the maximum value of the estimated action status is calculated. It is defined as a feature of This is because it is considered that a single feature can sufficiently describe a place where the function classification is a single function, and a feature with a low ratio is cut off as noise.

  Then, the moving object feature description unit 201 stores the calculated feature of the section in the moving object feature storage unit 202 (step S305).

  The above processing is repeated in the same manner for other day sections, time zones, and other sections.

  As described above, according to the present embodiment, the movement situation using the moving object and the action situation at the end point are estimated from the history of the position information of the user, and the estimated action situation at the end point of the movement area of the moving object is determined. By summing up, it is possible to select content that enhances the recency effect by using the result of the tabulation as a feature of the moving object, and the advertising effect can be efficiently enhanced.

  According to the present embodiment, when estimating the action situation at the movement section and the end point, the accompanying person's information is estimated and added, and when the estimated action situation is totaled, the estimated action situation of the accompanying person is calculated. If the estimated behavioral situation differs from the accompanying person, underestimating the estimated behavioral situation can eliminate the effects of estimation errors in the individual behavioral situation and accurately grasp the behavioral behavior of the group it can.

  According to the present embodiment, the behavioral situation at the start point of the movement section is also estimated, and when the estimated behavioral situation is totaled, by underestimating the one with a low transition probability of the estimated behavioral situation at the starting point-ending point, It is possible to reduce noise (outliers, etc.) of the estimated personal action situation.

  According to the present embodiment, when describing the feature of a moving object, the end point area of the moving section is described by referring to the functional section of the end point area of the moving section and describing the feature of the moving object according to the function classification. The characteristics of the moving object can be described according to.

DESCRIPTION OF SYMBOLS 100 ... Mobile body feature calculation part 101 ... Section selection part 102 ... Individual action condition correction | amendment part 103 ... Total part 104 ... Location feature utilization part 105 ... Location feature storage part 201 ... Mobile body feature description part 202 ... Mobile body feature storage part 300 ... Personal action status calculation unit 301 ... Location history storage unit 302 ... Personal location history acquisition unit 303 ... Movement mode determination unit 304 ... Movement route estimation unit 305 ... Movement section estimation unit 306 ... Personal behavior status estimation unit 307 ... Accompanying determination unit 308 ... Operating section position information storage unit 309 ... Personal action status history storage unit 310 ... Transition probability calculation unit

Claims (10)

Position history acquisition means for reading the history of the position information from the position history storage means for storing the history of the position information of the user;
A movement path estimation means for estimating the movement path of the user using the history of the position information;
A movement method determination means for determining a movement method of the user using a history of the position information;
When the movement method uses transportation, the travel section estimation means for estimating the travel section using the transportation by the user using the travel route;
An action situation estimating means for estimating an action situation at an end point of the moving section of the user using the history of the position information, and associating the action situation with the moving section and accumulating in a movement history accumulating means;
Wherein the movement history storage means, and behavioral state acquisition means for reading the behavioral state associated with the movement section corresponding to the write target section features moving body of the traffic institutions,
A counting means for counting the behavior status read by the behavior status acquisition means;
Description means for accumulating the aggregation results of the aggregation means as characteristics of the mobile object in the target section in the mobile object accumulation means;
A moving body feature description apparatus comprising: Detecting a companion accompanying the user using the history of the location information, and having a companion detection means for storing the companion information in the movement history storage means in association with the movement section;
When the companion information is stored in association with the movement section, the counting unit refers to the action situation associated with the movement section for the companion, and accompanies the user action situation and the accompaniment The mobile body feature description apparatus according to claim 1, wherein, when there is a difference in a person 's action situation, the action situation of the user and the action situation of the accompanying person are corrected to be the same . The behavior situation estimation means estimates the behavior situation at the start point of the movement section and accumulates it in the movement history accumulation means,
With reference to the movement history accumulating means, there is a transition probability calculating means for calculating a transition probability from the action situation at the start point to the action situation at the end point,
The moving body feature description apparatus according to claim 1, wherein the totaling unit excludes and summarizes the behavioral situation in which the transition probability is lower than a predetermined value.   The description means, when the functional section near the end point of the moving section is a single function, the largest of the totaling results is the feature of the moving body, and when the functional classification is multi-functional, the totaling result 4. The moving body feature description apparatus according to claim 1, wherein the moving body feature is described in accordance with the ratio of the moving body. A moving body feature description method executed by a computer,
Reading the history of the location information from the location history storage means for storing the location information history of the user;
Estimating the movement route of the user using the history of the location information;
Determining a movement method of the user using a history of the location information;
When the movement method uses transportation, the user estimates a movement section using the transportation using the movement route;
Estimating a behavior situation at the end point of the movement section of the user using the history of the location information, associating the behavior situation with the movement section, and accumulating in a movement history accumulation unit;
From the movement history storage means, reading out the behavioral state associated with the movement section corresponding to the write target section features moving body of the traffic institutions,
A step of counting the behavior status read in the step of reading the behavior status;
Accumulating the aggregation results of the aggregating step as characteristics of the moving object in the target section in a moving object feature storage unit;
A moving body feature description method characterized by comprising: Detecting a travel companion accompanying the user using the history of the location information, and storing the travel companion information in the travel history storage means in association with the travel section;
When the companion information is accumulated in association with the movement section, the step of counting refers to the action situation associated with the movement section for the companion, and the action situation of the user and the 6. The moving body feature description method according to claim 5, wherein when there is a difference in the behavior situation of the accompanying person, the behavior situation of the user and the behavior situation of the accompanying person are corrected to be the same and totaled. The step of estimating the behavior situation is to estimate the behavior situation at the start point of the movement section and accumulate it in the movement history accumulation means,
Referring to the movement history accumulating means, and calculating a transition probability from the action situation at the start point to the action situation at the end point;
7. The moving body feature description method according to claim 5 or 6, wherein the step of totaling excludes and aggregates the action situations whose transition probabilities are lower than a predetermined value.   In the step of accumulating the characteristics of the moving object, when the functional classification near the end point of the moving section is a single function, the largest of the total results is the characteristic of the moving object, and the functional classification is multifunctional. 8. The moving body feature description method according to claim 5, wherein, if there is, the characteristics of the moving body are described according to the ratio of the aggregation results.   A moving body feature description program that causes a computer to execute the moving body feature description method according to claim 5.   The recording medium which recorded the mobile body characteristic description program of Claim 9.

Images