CN117648556A

CN117648556A - Family membership identification method based on space-time big data

Info

Publication number: CN117648556A
Application number: CN202410118638.9A
Authority: CN
Inventors: 王鹰; 唐建中; 越海涛; 杨维; 杨凌霄; 林进聪; 蓝健财
Original assignee: Shenzhen Mastercom Technology Corp
Current assignee: Shenzhen Mastercom Technology Corp
Priority date: 2024-01-29
Filing date: 2024-01-29
Publication date: 2024-03-05
Anticipated expiration: 2044-01-29
Also published as: CN117648556B

Abstract

The application discloses a household membership identification method based on space-time big data. Obtaining user tracks of a preset number of users each day based on preset number of signaling data of the preset number of users each day, wherein the user tracks comprise dynamic user tracks and/or resident user tracks; determining the daily accompanying relation between any two users in the preset number of users according to the user track, wherein the accompanying relation comprises a dynamic accompanying relation and a resident accompanying relation; and identifying whether family membership exists between any two users based on the accompanying relation between any two users within a preset number of days. The user track is acquired based on the signaling data of the user, and the accompanying relation between any two users is determined based on the user track of the user, so that the family membership exists between the two users is determined, and the accuracy of identifying the family membership is effectively improved while the coverage breadth of the identification group is enlarged.

Description

Family membership identification method based on space-time big data

Technical Field

The application relates to the technical field of big data analysis, in particular to a method for identifying family membership based on space-time big data.

Background

In the conventional technology, a conversation record of a user or a friend relation of the user on a social platform is analyzed through a pre-trained identification model to identify family membership of the user, or the family membership is identified through using data of a broadband network; however, the training cost of the recognition model is high, and the recognition model is easy to interfere, for example, the conversation frequency is high or the net friend interaction frequency is high caused by work, so that the family membership recognition is inaccurate; however, the method of identifying family membership by using broadband usage data cannot avoid the situation that multiple families share broadband, resulting in low accuracy of identifying family membership.

The foregoing is merely provided to facilitate an understanding of the principles of the present application and is not admitted to be prior art.

Disclosure of Invention

The main purpose of the application is to provide a method for identifying family membership based on space-time big data, aiming at improving the accuracy of identifying family membership.

In order to achieve the above object, the present application provides a method for identifying family membership based on spatiotemporal big data, the method for identifying family membership based on spatiotemporal big data includes:

Obtaining user tracks of a preset number of users each day based on preset number of signaling data of the preset number of users each day, wherein the user tracks comprise dynamic user tracks and/or resident user tracks;

determining the daily accompanying relation between any two users in the preset number of users according to the user track, wherein the accompanying relation comprises a dynamic accompanying relation and a resident accompanying relation;

and identifying whether family membership exists between any two users based on the accompanying relation between any two users within a preset number of days.

Optionally, the step of obtaining the user track of the preset number of users each day based on the preset number of signaling data of the preset number of users each day includes:

acquiring signaling data of any user on the same day, and if the residence time of any user in any base station cell exceeds a preset residence threshold, outputting the signaling data of any user when the user resides in any base station cell as a residence state movement track of any user.

And acquiring signaling data of any user on the same day, and if the any user at least continuously passes through a preset number of base station cells, the residence time of each base station cell does not exceed a preset residence threshold, and the any user has similar motion directions in the continuously passed preset number of base station cells, outputting the signaling data of the any user when the any user passes through the base station cells as a motion track of the any user.

Optionally, the track information of the resident user track includes a resident time period, a resident point position and a resident cell; the step of determining the daily accompanying relationship between any two users in the preset number of users according to the user track comprises the following steps:

for the first resident state user track of one user and the second resident state user track of the other user in any two users, if the first resident state user track and the second resident state user track of the other user simultaneously meet the following conditions:

the first resident point position on the first resident state user track and the second resident point position on the second resident state user track have an intersection in a resident time period, and the intersection time length with the intersection exceeds a preset overlapping time length threshold;

the distance between the first standing point position and the second standing point position is smaller than a preset standing point distance threshold;

The residence cells of the first residence point position and the second residence point position have an intersection, and the sum of residence time lengths of residence cells with the intersection exceeds a preset residence total time length threshold;

it is determined that there is a resident state companion relationship between any two users on the same day.

Optionally, the step of determining the daily concomitant relationship between any two users in the preset number of users according to the user track includes:

based on the dynamic user track of any two users on the same day, track OD detection and/or space-time accompanying detection are carried out, so that a detection result is obtained;

and determining whether a dynamic accompanying relation exists between any two users on the same day based on the detection result.

Optionally, the track information of the dynamic user track includes a track time stamp, a path section and a track total distance; the step of carrying out space-time accompanying detection based on the dynamic user track of any two users on the same day to obtain a detection result comprises the following steps:

determining a coincident road section based on the road sections of any two users on the same day;

determining the moment when any two users arrive and/or leave the coincident road section based on the track time stamp on the coincident road section of any two users on the same day;

If the time difference of any two users arriving at and/or leaving the overlapped road section is smaller than a preset difference threshold, determining that the overlapped road section is a short-distance accompanying road section;

counting the total accompanying distance of the short-distance accompanying road sections on the dynamic user track of any two users on the same day;

calculating to obtain an accompanying distance ratio based on the accompanying total distance and the track total distance of the dynamic user tracks of any two users on the same day;

and if the accompanying total distance is larger than a preset total distance threshold and the accompanying distance ratio is larger than the preset accompanying distance threshold, determining that the detection result is that any two users have space-time accompanying.

Optionally, the step of determining whether a dynamic accompanying relationship exists between any two users on the same day based on the detection result includes:

if the detection result is that the tracks OD of any two users are the same and space-time accompaniment exists for any two users, the same accompanying relationship between any two users on the same day is determined to be a dynamic accompanying relationship.

Optionally, the step of identifying whether the family membership exists between any two users based on the accompanying relationship between any two users within a preset number of days includes:

counting the first days of the resident state accompanying relation as the accompanying relation between any two users in a preset night period within a preset number of days;

Counting other time periods except for a preset night time period in one day between any two users in a preset number of days, and simultaneously, counting a second number of days in which the dynamic accompanying relation and the resident accompanying relation exist;

and if the ratio of the first days to the preset days is larger than a preset first ratio threshold and the ratio of the second days to the preset days is larger than a preset second ratio threshold, determining that family membership exists between any two users.

Optionally, the method further comprises:

outputting the users with the family membership to a user list to be clustered;

traversing the user list to be clustered in sequence;

if the family membership does not exist between the current traversal user and the traversed user, a family group of the current traversal user is newly built;

and if the family membership exists between the current traversed user and the traversed user, adding the current traversed user into the family group of the traversed user.

Optionally, the method further comprises:

and if the current traversed user has family membership with one or more traversed users in at least a first family group and one or more traversed users in a second family group, combining at least the first family group and the second family group into a new family group, and adding the current traversed user into the new family group.

The method for identifying family membership based on space-time big data comprises the following steps: obtaining user tracks of a preset number of users each day based on preset number of signaling data of the preset number of users each day, wherein the user tracks comprise dynamic user tracks and/or resident user tracks; determining the daily accompanying relation between any two users in the preset number of users according to the user track, wherein the accompanying relation comprises a dynamic accompanying relation and a resident accompanying relation; and identifying whether family membership exists between any two users based on the accompanying relation between any two users within a preset number of days.

In the conventional technology, the family membership is usually identified through a pre-trained identification model, or the user sharing the same broadband for a long time is judged to be the family member through the use data of the broadband; the method for judging family membership through the shared broadband ignores the situation that the broadband is not used and the shared broadband such as a co-renting and dormitory is not used, the identification method is single, the coverage crowd is not comprehensive, the identification effect is poor, and the accuracy is low.

Different from the conventional technology, the household membership identification method based on the space-time big data obtains the user tracks of the preset number of users based on the massive signaling data provided by operators, ensures the accuracy of data sources, avoids the problem of incomplete coverage crowd, generates instant signaling data as long as the intelligent terminal carried by the users communicates with other devices or networks, and can distinguish different users according to the unique user identification in the user signaling data even if sharing the same broadband, thereby effectively expanding coverage crowd breadth; judging the user track intersection condition between any two users through the user tracks, dividing the identified user tracks into a dynamic user track and a resident user track, determining a dynamic accompanying relation between the two users according to the dynamic user tracks of any two users, determining the resident accompanying relation between the two users according to any two users, and finally identifying whether family membership exists between any two users based on the accompanying relation between any two users; thus, the coverage group breadth is enlarged by processing the signaling data of the user; meanwhile, on the basis of acquiring the user track, the accompanying data between any two users are acquired, so that whether the family membership exists between any two users is determined, training of a model is not needed, the cost of identifying the family membership is effectively reduced, and the accuracy of identifying the family membership is improved.

Drawings

Fig. 1 is a schematic structural diagram of a family membership identification device of a hardware running environment according to an embodiment of the present application;

FIG. 2 is a schematic flow chart of a first embodiment of the present application;

FIG. 3 is a schematic view of user clustering according to an embodiment of the present application;

fig. 4 is a schematic view of family group clustering related to an embodiment of the present application;

fig. 5 is a schematic functional block diagram of a family membership identification device according to an embodiment of the present application.

The realization, functional characteristics and advantages of the present application will be further described with reference to the embodiments, referring to the attached drawings.

Detailed Description

It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

Referring to fig. 1, fig. 1 is a schematic diagram of a family membership identification device of a hardware running environment according to an embodiment of the present application.

As shown in fig. 1, the family membership identification device may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a WIreless interface (e.g., a WIreless-FIdelity (WI-FI) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) Memory or a stable nonvolatile Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.

It will be appreciated by those skilled in the art that the structure shown in fig. 1 does not constitute a limitation of the family membership identification device, and may include more or fewer components than shown, or may combine certain components, or a different arrangement of components.

As shown in fig. 1, an operating system, a network communication module, a user interface module, and a computer program may be included in the memory 1005 as one type of storage medium.

In the family membership identification device shown in fig. 1, the network interface 1004 is mainly used for data communication with other devices; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the family membership identification device may be provided in the family membership identification device, where the family membership identification device invokes a computer program stored in the memory 1005 through the processor 1001, and executes the method for identifying family membership based on spatiotemporal big data provided in the embodiment of the present application:

In an embodiment, the processor 1001 may call a computer program stored in the memory 1005, and further perform the following operations:

the step of obtaining the user track of the preset number of users each day based on the preset number of signaling data of the preset number of users each day includes:

the track information of the resident user track comprises a resident time period, a resident point position and a resident cell; the step of determining the daily accompanying relationship between any two users in the preset number of users according to the user track comprises the following steps:

the step of determining the daily accompanying relationship between any two users in the preset number of users according to the user track comprises the following steps:

the track information of the dynamic user track comprises a track time stamp, a path section and a track total distance; the step of carrying out space-time accompanying detection based on the dynamic user track of any two users on the same day to obtain a detection result comprises the following steps:

the step of determining whether a dynamic accompanying relationship exists between any two users on the same day based on the detection result comprises the following steps:

the step of identifying whether family membership exists between any two users based on the accompanying relationship between any two users within a preset number of days comprises the following steps:

the method further comprises the steps of:

outputting the users with the family membership to a user list to be clustered;

traversing the user list to be clustered in sequence;

the method further comprises the steps of:

Referring to fig. 2, fig. 2 is a flowchart of a first embodiment of a method for identifying family membership based on spatio-temporal big data.

For the communication industry, identifying family members is an important premise for developing accurate marketing and customer management, however, the existing schemes for identifying family members often have the defects of single identification mode and low identification accuracy.

Based on massive space-time big data, a travel mode and a rule of a city or an area are mined, the accompanying travel characteristics of the crowd are identified, and the method is a research hotspot of the space-time big data. The 4, 5G communication network operated by the communication carrier has all users and all time signaling data, and these signaling data have location tags, which are typical space-time large data. According to the household membership identification method based on the space-time big data, the user track of a single user on the same day is firstly determined based on the space-time big data of operators, then the user is subjected to the accompanying relation analysis based on the user tracks of a plurality of users on the same day, and then the household membership is identified, so that the household membership identification method has the advantages of being wide in coverage crowd, low in implementation cost, high in accuracy and the like.

In this embodiment, the method for identifying family membership based on spatiotemporal big data includes:

step S10: obtaining user tracks of a preset number of users each day based on preset number of signaling data of the preset number of users each day, wherein the user tracks comprise dynamic user tracks and/or resident user tracks;

in this embodiment, signaling data of users are obtained through 4G and 5G communication networks operated by a communication operator, and user trajectories of the users in the preset number are obtained based on the preset number of signaling data of the users in the preset number each day, where the user trajectories include a dynamic user trajectory and a resident user trajectory; the preset number can be adjusted according to actual conditions, for example, the preset number of users can be all users, or some users in all users, and the preset number of signaling data can refer to signaling data corresponding to the preset number of users;

The user track can be represented as a track map, track data and the like, the dynamic user track is a track moving in a large range in the user track, for example, in the user track of a certain user on the same day, the situation that the residence time of the user in a certain period of time/a certain track section is less than the residence threshold, the user track continuously passes through a plurality of base station cells and has obvious movement direction is shown, and the track of the user in the period of time/the track section is the dynamic user track; the resident state user track is a track where a user stays for a long time in the user track, for example, in the user track of a certain user on the same day, when the time length of stay of the user in a certain base station cell exceeds a resident threshold in a certain period of time/in a certain track, the user is in a resident state in the base station cell, and the user in the period of time/the track section track is the resident state user track.

In an embodiment, the dynamic user track and the resident user track may be output in a form of a data table, where the resident data table corresponding to the resident user track includes at least a user identity, a base station cell identity, a resident start time, a resident end time, and a resident point position; the dynamic data table corresponding to the dynamic user track at least comprises a user identity, a base station cell identity, a dynamic starting time, a dynamic ending time, a dynamic starting position, a dynamic arrival position, a total distance of the motion track, a passing road section and a road section arrival time.

The resident state data table refers to table 1:

TABLE 1

The kinematic data table refers to table 2:

TABLE 2

It can be understood that, for the whole user track or the dynamic user track in the user track, road fitting can be further performed according to the information in the dynamic data table, so that the user track or the motion track of the user on the road on a certain day is fitted, and at present, the scheme of constructing the road network track of the user by utilizing massive signaling data is mature in the industry and will not be described herein.

The preset residence threshold can be adjusted according to practical situations, for example, the preset residence threshold is set to be 1 hour; and reading signaling data of any user, if the residence time of the user in a certain base station cell exceeds a set residence threshold, determining that the user is in a residence state in the base station cell, extracting the signaling data of the user in the residence state period on the current day by setting a proper preset residence threshold, and outputting the signaling data of the user in the base station cell as a residence state user track of the user.

The preset number is required to be greater than 1, and the preset number and the preset residence threshold can be adjusted according to practical situations, for example, the preset number is set to be 2, and the residence threshold is set to be 1 hour; the motion track or the velocity vector in the motion state of the user has the same or similar direction, and the motion direction may be a motion direction obtained based on signaling data of the user, or may be a motion direction obtained based on an installation position of a base station cell, for example, a motion direction of a third base station cell relative to a second base station cell and a motion direction of the second base station cell relative to a first base station cell are the same or similar in three base station cells through which the user passes. If the user continuously passes through the preset number of base station cells, the residence time in the passing base station cells does not exceed the set residence threshold and has obvious movement direction, determining that the user is in movement state when passing through the base station cells, and outputting the signaling data when the user passes through the base station cells as the movement state user track of the user.

Preferably, the subsequent step is implemented based on a preset number of signaling data of a preset number of users each day, and the signaling data may be less duration, such as half a day, signaling data within a certain period of time, or longer duration, such as two days, signaling data, where in this embodiment, the acquisition duration of the user signaling data for data processing in the method for identifying family membership based on spatio-temporal big data is not limited.

The signaling data records signaling interaction information between an intelligent terminal and a base station in a mobile communication network, and the intelligent terminal comprises, but is not limited to, a mobile phone, a bracelet and an intelligent watch; the signaling data includes, but is not limited to, base station Cell Identity ECI/NCI, ECI (E-UTRANCell Identifier, E-UTRAN Cell unique Identity) is a unique identifier for identifying a 4G Cell, NCI (NR Cell Identity) is a unique identifier for identifying a 5G Cell, user Identity (already desensitized), time stamp (accurate to millisecond), longitude/latitude.

Step S20: determining the daily accompanying relation between any two users in the preset number of users according to the user track, wherein the accompanying relation comprises a dynamic accompanying relation and a resident accompanying relation;

The dynamic accompanying relation refers to that an intersection exists in a dynamic user track of any two users, for example, in a dynamic user track of a certain day of a certain two users, the two users are found to start from the same starting point or start similarly at the same moment, pass through the same or all the same road sections and reach the same end or end similarly, and the existence of the dynamic accompanying relation between the two users is indicated; the resident state accompanying relation refers to that an intersection exists in resident state user tracks of any two users, for example, in resident state user tracks of a certain day of a certain two users, if the two users reside in the same time period and the same position, the resident state accompanying relation exists between the two users;

in addition, ways to determine the concomitant relationship include, but are not limited to, comparing and matching user trajectories. In an alternative embodiment, determining the accompanying characteristics between any two users among the preset number of users, and determining the accompanying relation between any two users according to the accompanying characteristics; the accompanying features may be parameters of part or all of the user tracks, and the accompanying relationship between any two users may be determined according to the similarity of the accompanying features.

Step S30: and identifying whether family membership exists between any two users based on the accompanying relation between any two users within a preset number of days.

In this embodiment, to ensure accuracy of data analysis, whether a family membership exists between two users is identified based on a concomitant relationship between any two users for a plurality of consecutive days, the preset number of days may be one month, and the preset number of days may be adjusted according to actual conditions;

wherein the manner of identifying whether there is family membership includes: counting the total number of occurrence times of the dynamic accompanying relation and the resident accompanying relation of two users in preset days, judging whether the total number of occurrence times exceeds a set number threshold, and determining that the two users are family membership by assuming that the preset days are one month (30 days), the number threshold is 40 times and the number of occurrence times of the accompanying relation of the two users is 15 times, the resident accompanying relation is 34 times and the total number of occurrence times exceeds the set number threshold in one month; or, after counting the number of times that the dynamic accompanying relationship and the resident accompanying relationship of any two users respectively appear in a preset number of days, respectively judging whether the number of times exceeds the respective set number of times threshold, and determining that the two users are family membership when both the number of times exceeds the number of times threshold.

Preferably, in a second embodiment, based on the first embodiment, the track information of the resident user track includes a resident time period, a resident location, and a resident cell; the step of determining the daily accompanying relationship between any two users in the preset number of users according to the user track comprises the following steps:

In an embodiment, the residence time period can be calculated based on residence starting time and residence ending time, and the residence point position of the user can be obtained according to longitude and latitude information in signaling data of the user; for the resident state user track of any two users, if the resident state user track simultaneously satisfies: the method comprises the steps that a first resident point position on a first resident state user track of one user and a second resident point position on a second resident state user track of the other user have an intersection in a resident time period, and the intersection time length of the intersection exceeds a preset overlapping time length threshold; the distance between the first resident point position and the second resident point position is smaller than a preset resident point distance threshold, specifically, the resident point distance is calculated by adopting a seminormal distance formula based on longitude and latitude information of the user resident point position; and determining that the resident state accompanying relation is in the resident time period exceeding the preset overlapping time length threshold in the current accompanying relation between any two users, and screening the three judgment layers to ensure the identification accuracy, so that the algorithm is simple without training, and the cost is effectively reduced.

In one embodiment, for a first resident state user track of one user and a second resident state user track of the other user in any two users, if both: the first resident point position on the first resident state user track and the second resident point position on the second resident state user track have an intersection in a resident time period, and the intersection time length with the intersection exceeds a preset overlapping time length threshold; the distance between the first standing point position and the second standing point position is smaller than a preset standing point distance threshold, the residence cells of the first standing point position and the second standing point position have an intersection, and the sum of residence time lengths of residence cells with the intersection exceeds a preset residence total time length threshold; and if any one or two judgment conditions exist, determining that the resident state accompanying relation exists between any two users on the same day.

Assuming that any two users are a user A and a user B, a preset overlapping time threshold is 1 hour, in a resident state user track of the user A and the user B on the same day, from the zero point in the morning to 8 points in the morning, the intersection time of the user A and the user B is in a resident state, and exceeds the preset resident overlapping time, which indicates that the user A and the user B have accompanying relations in a resident time period, but only judging that the error of the resident time period is too large, so that whether the positions of the user A and the user B in the resident time period with the intersection are identical or not is also required to be judged, a position error range can be set according to actual conditions to ensure the accuracy of identification, if the distance between the resident point positions of the user A and the user B in the resident time period with the intersection does not exceed the position error range, the positions of the user A and the user B are similar or identical, and the user A and the user B have the accompanying relations in the resident time period and the resident point positions; in a mobile communication network, the coverage area of a base station cell is limited, the overlapping situation of the base station cell inevitably occurs, if the resident cells of the user A and the user B have an intersection, the resident cells with the intersection comprise but are not limited to the same cell or different cells with overlapping ranges, and the sum of the duration of the resident time periods of the user A and the user B overlapped in the resident cells with the intersection exceeds a preset resident total duration threshold, the resident state accompanying relation between the user A and the user B is determined.

Preferably, in a third embodiment, based on the first embodiment, the step of determining, according to the user track, a daily concomitant relationship between any two users in a preset number of users includes:

In one embodiment, the user's track OD (Origin to Destination, start to end) is used to describe the relationship between the user's movement track and start/end over a period of time, and may be acquired and analyzed in a variety of ways, such as GPS track data, movement positioning data, public transportation swipe data, etc. If the tracks OD of any two users are the same, the fact that the two users select the same starting point and end point to travel in a certain period of time is indicated that the two users with the same track OD are likely to travel together at the same time; the space-time accompaniment is used for describing the association relation between any two user tracks in a certain time and space range; whether the user tracks OD of any two users are the same or whether space-time accompaniment exists between any two users is detected, and the detection method can be independently used as a basis for judging the dynamic accompaniment relationship between any two users; preferably, the family member identification method combines detection of the user track OD and detection of the user space-time accompanying condition, ensures reliability of detection results, and improves accuracy of family member relationship identification.

Assuming that any two users are a user A and a user B respectively, the dynamic user track on the same day is a track A and a track B respectively, and the track OD detection step comprises the following steps: if the absolute value of the difference between the dynamic start time and the dynamic end time of the track a and the track B is smaller than a preset time threshold (for example, 5 minutes), and the distance between the dynamic start points and the distance between the dynamic arrival points are smaller than a preset distance threshold (for example, 500 meters), it is indicated that the tracks OD of the two users are the same, otherwise, the tracks OD of the two users are different.

Preferably, in a fourth embodiment, based on the above third embodiment, the track information of the dynamic user track includes a track time stamp, a route section, and a total track distance; the step of carrying out space-time accompanying detection based on the dynamic user track of any two users on the same day to obtain a detection result comprises the following steps:

In one embodiment, the step of performing space-time detection on any two users includes: determining a coincident road section in the motion trail of two users based on the route road section of any two users on the same day, wherein the coincident road section refers to the road section of which all or most of the users pass through in the travel process, and the motion trail of the two users can be fitted according to the signaling data of the two users; after determining the coincident road sections in the tracks of two users, determining the moment when any two users arrive and/or leave the coincident road sections based on the track time stamp of the two users on the coincident road sections, wherein the intelligent terminal carried by the user can continuously communicate with the base station, and the signaling data generated in the communication process comprises the data such as the communication place, the communication moment and the like of the intelligent terminal; the track time stamp is the time information corresponding to each recorded standing point position of the user in the signaling data.

If a coincident road section exists in the user tracks of the two users, and the time difference of the two users arriving at and/or leaving from the coincident road section is smaller than a preset difference threshold, determining the coincident road section as a short-distance accompanying road section, wherein the two users have accompanying relations on the coincident road section, counting the accompanying total distance of the short-distance accompanying road section on the dynamic user track of the two users on the same day, and calculating an accompanying distance ratio based on the accompanying total distance and the track total distance of the dynamic user track of any two users on the same day, wherein the accompanying distance ratio is calculated by referring to the following formula:

in the method, in the process of the invention,the track A distance and the track B distance represent the total distance of the user tracks of the two users; if the total accompanying distance is larger than the preset total distance threshold and the accompanying distance ratio is larger than the preset accompanying distance threshold, determining that the detection result is that space-time accompanying exists in the dynamic user tracks of any two users, and detecting the dynamic user tracks of the two users in a space-time manner based on the space-time relationship between the two usersWhether a dynamic accompanying relation exists between the two users or not is determined according to the space-time accompanying detection condition of the two users, when the space-time accompanying exists between the two users, the dynamic accompanying relation exists between the two users in the user track of the same day, and then whether family membership exists between the users or not is identified based on the accompanying relation between the two users, so that identification accuracy is effectively improved. The preset difference threshold, the accompanying distance threshold and the preset total distance threshold may be adjusted according to practical situations, for example, the preset difference threshold may be set to 2 minutes, the preset total distance threshold may be set to 2000 meters, and the accompanying distance threshold may be set to 0.7.

Preferably, in a fifth embodiment, based on the fourth embodiment, the step of determining whether a dynamic accompanying relationship exists between any two users on the same day based on the detection result includes:

In an embodiment, if the detection result is that the tracks OD of any two users are the same and space-time accompaniment exists for any two users, the same day accompaniment relationship between any two users is determined to be a dynamic accompaniment relationship, and meanwhile, track OD detection and space-time accompaniment detection are performed on the dynamic user tracks of the two users, so that the reliability of the detection result is ensured, and the accuracy of identifying family membership is improved.

Preferably, in a sixth embodiment, based on the first embodiment, the step of identifying whether there is a family membership between any two users based on the concomitant relationship between any two users within a preset number of days includes:

In an embodiment, counting the first days of the resident state accompanying relationship between any two users in the preset night time period within the continuous preset days, wherein the preset days and the preset night time period can be adjusted according to actual conditions, for example, the preset days are set to be 30 days, and the preset night time period is set to be 22 pm to 7 pm of the second day; counting the second days in which the resident state accompanying relation and the dynamic state accompanying relation exist in other time periods except for the preset night time period in at least one day between any two users in the preset days; if the ratio of the first days to the preset days is greater than the preset first ratio threshold, and the ratio of the second days to the preset days is greater than the preset second ratio threshold, determining that the family membership exists between any two users, wherein the preset first ratio threshold and the preset second ratio threshold can be adjusted according to actual conditions, for example, are respectively set to 0.6 and 0.2. The recognition algorithm for determining whether any two users have family membership is as follows:

1) In the data of m consecutive days, there is a night period residence state accompanying relation exceeding n days, and n/m > threshold beta (such as 0.6);

2) In data for m consecutive days, there is more than one day while there are other period resident state accompanying relations and dynamic state accompanying relations, and n/m > threshold θ (e.g., 0.2).

If two users meet the two conditions at the same time, the two users are indicated to have family membership.

In the travel tracks among family members, the family members usually rest at home in the night period, which can lead the tracks to show the characteristic of stay state at night, namely a plurality of users stay at the same place and are in stay state accompanying relation with each other; on weekends or other times, family members may go out to perform activities together, such as shopping together, playing together, etc., which may cause their trajectories to show obvious motion states in the period, that is, multiple users move together in the same period, are in motion state accompanying relations with each other, identify whether family membership exists between any two users based on the accompanying relations for multiple consecutive days, ensure accuracy of identification results, and meanwhile, by analyzing resident state accompanying relations and motion state accompanying relations of users for multiple consecutive days, an operator may also know interaction and behavior patterns between family members, so that information required by users is accurately pushed to users, marketing is accurate, and personalized family services are provided for travel planning, family activity arrangement, etc.

Preferably, in a seventh embodiment, based on the first embodiment, the method further includes:

outputting the users with the family membership to a user list to be clustered;

traversing the user list to be clustered in sequence;

Preferably, when the users are output to the user list to be clustered, the users can be at least sorted according to the number of family membership of the users, and the clustering efficiency is effectively improved. Referring to fig. 3, fig. 3 is a family membership diagram of users a/B/C/D/E, where a connection line in the diagram represents a directly identified family membership existing between users, where the number of identified family membership of users a/B/C/D/E is 2, 3, 2 and 1, respectively, and the ranking after being output to the user list to be clustered may be user C/a/B/D/E.

Considering that 3 or more family members exist in most families, such as a typical three-family, parents and children, all members of the same family need to be gathered into one family group on the basis of identifying family membership among users, and assistance is provided for developing a family market. In a family, there may be a direct-identification family membership between any two members, or there may be only a part of members with a direct-identification family membership, for example, there may be a direct-identification family membership between mom and child, between mom and father, but there may be a small chance that child and father go out together at ordinary times, and the standard of the direct-identification family membership may not be reached, at this time, the real family membership of child and father may not be identified, and by traversing a user list with family membership, users with direct or indirect family membership are clustered into the same family group, so that users with direct family membership in the same family group are not detected, and are determined to be family membership, thereby effectively improving the accuracy of family membership identification.

Preferably, in an eighth embodiment, based on the seventh embodiment, the method further includes:

In an embodiment, if the currently traversed user has direct identification of family membership with members of the existing 2 or more family groups, the family groups are first combined into a new large family group, the current user is also combined into the combined family group, and the family membership data of each user in the family group is updated by combining the family groups, so that the accuracy of family membership identification is ensured.

Referring to fig. 4, if the user F has family membership with the user B in the family group 1 and the user E in the family group 2, respectively, the family group 1 and the family group 2 are combined into a new family group, and the user F is added to the new family group, thereby obtaining a family group 3.

In addition, for the user with the family membership of 0, after the traverse is finished, a family group is respectively established for each user with the family membership of 0, and the family groups of all the users with the family membership and the family group with the family membership of 0 are output as a family membership table.

Further, referring to fig. 5, the present application also provides a family membership identification device, including:

the track module M1 is used for obtaining the user tracks of the preset number of users each day based on the preset number of signaling data of the preset number of users each day, wherein the user tracks comprise dynamic user tracks and/or resident user tracks;

the accompanying module M2 is used for determining the daily accompanying relation between any two users in the preset number of users according to the user track, wherein the accompanying relation comprises a dynamic accompanying relation and a resident accompanying relation;

the identifying module M3 is configured to identify whether a family membership exists between any two users based on the concomitant relationship between any two users within a preset number of days.

The family membership identification device provided by the application adopts the family membership identification method based on the space-time big data in the embodiment, and aims to improve the accuracy of family membership identification. Compared with the conventional technology, the beneficial effects of the family membership identification device provided by the embodiment of the application are the same as those of the family membership identification method based on space-time big data provided by the embodiment, and other technical features in the family membership identification device are the same as those disclosed by the method of the embodiment, so that details are not repeated.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.

The foregoing embodiment numbers of the present application are merely for describing, and do not represent advantages or disadvantages of the embodiments.

From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) as described above, including several instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method described in the embodiments of the present application.

The foregoing description is only of the preferred embodiments of the present application, and is not intended to limit the scope of the claims, and all equivalent structures or equivalent processes using the descriptions and drawings of the present application, or direct or indirect application in other related technical fields are included in the scope of the claims of the present application.

Claims

1. A method for identifying family membership based on spatiotemporal big data, characterized in that the method for identifying family membership based on spatiotemporal big data comprises the following steps:

2. The method for recognizing family membership based on spatiotemporal big data according to claim 1, wherein the step of obtaining a user trajectory of a preset number of users per day based on a preset number of signaling data of the preset number of users per day comprises:

3. The method for recognizing family membership based on spatiotemporal big data according to claim 1, wherein the step of obtaining a user trajectory of a preset number of users per day based on a preset number of signaling data of the preset number of users per day comprises:

4. The method for identifying family membership based on spatiotemporal big data according to claim 1, wherein the track information of the resident user track includes resident time period, resident point position, resident cell; the step of determining the daily accompanying relationship between any two users in the preset number of users according to the user track comprises the following steps:

5. The method for recognizing family membership based on spatiotemporal big data according to claim 1, wherein the step of determining a daily concomitant relationship between any two users among a preset number of users according to the user trajectory comprises:

6. The method for recognizing family membership based on spatiotemporal big data according to claim 5, wherein the track information of the dynamic user track includes a track time stamp, a route section, a track total distance; the step of carrying out space-time accompanying detection based on the dynamic user track of any two users on the same day to obtain a detection result comprises the following steps:

7. The method for recognizing family membership based on spatiotemporal big data according to claim 6, wherein the step of determining whether a dynamic accompanying relationship exists between any two users on the same day based on the detection result comprises:

8. The method for recognizing family membership based on spatiotemporal big data according to claim 1, wherein the step of recognizing whether family membership exists between any two users based on the concomitant relationship between any two users within a preset number of days comprises:

9. The method of spatiotemporal big data based family membership identification of claim 1, further comprising:

outputting the users with the family membership to a user list to be clustered;

traversing the user list to be clustered in sequence;

10. The method of spatiotemporal big data based family membership identification of claim 9, further comprising: