CN110378002B - Social relationship modeling method based on movement track - Google Patents
Social relationship modeling method based on movement track Download PDFInfo
- Publication number
- CN110378002B CN110378002B CN201910625340.6A CN201910625340A CN110378002B CN 110378002 B CN110378002 B CN 110378002B CN 201910625340 A CN201910625340 A CN 201910625340A CN 110378002 B CN110378002 B CN 110378002B
- Authority
- CN
- China
- Prior art keywords
- user
- meeting
- time
- track
- users
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Abstract
The invention discloses a social relation modeling method based on a movement track, which comprises the following steps: 1) Recording the track points of the user according to the collected track data of the user in a set time period at preset time intervals to form a time stamp sequence of the track of the user; 2) Judging whether the users meet according to the timestamp sequence of each user track, if yes, recording meeting record vectors between every two users; 3) Constructing a social relation model based on the moving track according to the meeting record vector; 4) And calculating the social relationship strength between the users according to the social relationship model, and dividing the mobile communities to which the users belong. The invention constructs a calculation method for measuring social relationship strength by calculating the meeting times and meeting time among different mobile users, and the method converts the social relationship strength into a weighted social relationship network, thereby constructing different user communities and providing assistance for more accurate location service based on tracks.
Description
Technical Field
The invention relates to a social network technology, in particular to a social relationship modeling method based on a movement track.
Background
As portable devices and positioning technologies mature, the amount of trajectory data is increasing. The large-scale track data characterizes the space-time dynamics of individuals and groups, contains the behavior information of human beings, vehicles and animals, and how to extract potential and meaningful knowledge from the behavior information becomes a new research hot spot and a new problem in the field of data mining. The mobile user position prediction is based on the premise of the position service, the hidden social relationship of the mobile user in the track data is potential useful information, and the accuracy of the mobile mode mining and the mobile position prediction can be improved. How to measure the strength of social relationships between people in a social network has been a difficult problem in social network relationship analysis. The information such as interaction frequency, time, position, distance, track similarity and the like can directly reflect the interaction relationship and relationship strength among people, and through analysis processing of the information, measurement indexes which take three aspects of interaction frequency, connection times and intimacy degree as measurement relationship strength are formed, and social relationship can be deduced by using the simultaneous occurrence times of different users in the same geographic position. Through analysis of the social psychology related research results, the relationship strength among people is considered to be closely related to the track similarity among people and the similarity of daily behaviors. The strength of the relationship between people is generally calculated from the two aspects described above.
In the existing track prediction method, the preliminary position prediction is generally performed on the nodes based on a Markov model, the prediction result is corrected by combining the daily movement law and the social relationship, the daily life can be divided into different time periods, the track position of each event period is predicted by using the Markov model, and then the social relationship is used for correction. When modeling the social relationship, the degree of tightness of the relationship between the nodes is measured mainly by calculating the total number of times of meeting between any two mobile nodes to form a relationship network, and then the communities contained in the relationship network are calculated by using a complex network community discovery algorithm, wherein the nodes of the same community have stronger social relationship, and the nodes of different communities have weaker social relationship. The method has a certain limitation, firstly, because the relationship strength between the mobile nodes in reality is related to the meeting times and the time length of each meeting, and the influence of selecting different complex network community discovery algorithms on the result is large, the construction of a proper social relationship model and community discovery algorithm is a problem worthy of research.
Disclosure of Invention
The invention aims to solve the technical problem of providing a social relationship modeling method based on a moving track aiming at the defects in the prior art.
The technical scheme adopted for solving the technical problems is as follows: a social relation modeling method based on a movement track comprises the following steps:
1) Recording the track points of the user according to the collected track data of the user in a set time period at preset time intervals to form a time stamp sequence of the track of the user;
the track data of the user is expressed as%<l 1 ,t 1 ,a 1 >,…,<l i ,t i ,a i >，…，<l n ，t n ，a n >Wherein i is more than or equal to 1 and less than or equal to n, l i ＝＜x i ,y i Longitude and latitude data representing the location of the user, t i For user arrival l i Time, a i The motion attribute of the corresponding moment of the track data comprises a direction and a speed; and satisfy t i ＜t i+1 ；
2) Judging whether the users meet according to the timestamp sequence of each user track, if yes, recording meeting record vector e between every two users ij ＝(v j ,d ij ,n ij ,t ij ) The vector records the user v i And user v j Wherein v is j For user v i Is t ij For user v i And user v j N, n ij For user v i And user v j D is the number of times of meeting ij For user v i And user v j Is the meeting date of (2);
the user v i And user v j The manner of judging the meeting is as follows:
for user v i And user v j Intercepting the coincident part of the time stamp sequence within the same starting time and ending time range, and the user v i At t 1 The position of the time is point b (la 1 ,lo 1 ) (la is longitude and lo is latitude), at t 2 Is the position of point c (la 3 ,lo 3 ) User v j The position at time t is point a (la 2 ,lo 2 ) And satisfy t 1 ＜t＜t 2 Calculating the Euclidean distance L between a and b and between a and c 1 ，L 2 The method comprises the steps of carrying out a first treatment on the surface of the Setting the distance threshold value as theta D If L 1 ≤θ D And L is 2 ≤θ D Simultaneously calculating the time length delta t of meeting of two users, wherein delta t=t 2 -t 1 Setting the minimum meeting time threshold value as theta T If Deltat is greater than or equal to theta T User v i And user v j If not, the two parts are not met;
3) Constructing a social relation model based on the moving track according to the meeting record vector;
from the encounter record vector e= [ E 1 ,e 2 ,…,e n ]The meeting matrix of the user nodes is obtained as follows:
based on the meeting matrix, n is counted ij Component sum t ij The components, calculate the meeting times and the total meeting time of each user node and all other user nodes, get the meeting times matrix k [ i ] of the nodes][j]And an encounter time matrix ti][j]；
The conversion of the meeting times matrix k [ i ] [ j ] into the meeting times probability matrix is as follows:
the probability matrix of the number of encounters between the mobile nodes is as follows:
according to the meeting time t [ i ] [ j ], the meeting time probability matrix among the mobile nodes is calculated as follows:
obtaining the meeting time probability matrix among the mobile nodes according to the above method:
establishing a social relation matrix P SR The following are provided:
P SR ＝(1-ξ)P T +ξ*P N (6)
wherein, xi is an adjusting factor which is more than or equal to 0 and less than 1 and is used for adjusting the meeting frequency matrix P N And an encounter time matrix P T Influence on the social relationship strength measurement between the mobile nodes;
4) And calculating the social relationship strength between the users according to the social relationship model, and dividing the mobile communities to which the users belong.
According to the above scheme, step 2) judges whether users meet according to the time stamp sequence of the track of each user, and first, pre-processes the time stamp sequence of the track of each user, wherein the pre-processes are as follows: if the track time stamp sequences of the users A and B overlap in a certain time period, a is set 1 、a 2 For user A at
Position of time, b 1 、b 2 For user B +.>
The position of the moment, the original track data is preprocessed in the overlapping time period, if +.>
And is also provided with
Will a 1 、a 2 Merging into one time point; after processing the entire time period, both sequences of the user A, B are traversed simultaneously in time-stamped order.
The invention has the beneficial effects that: the invention constructs a calculation method for measuring social relationship strength by calculating the meeting times and meeting time among different mobile users, the method converts the social relationship strength into a weighted social relationship network, and the network intended by the user can be obtained by changing input parameters, such as defining different meeting time lengths, thereby constructing different user communities and providing assistance for more accurate location service based on tracks.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a flow chart of a method of an embodiment of the present invention;
FIG. 2 is a diagram of the meeting of users A and B according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of overlapping time stamp sequences of node A and node B according to an embodiment of the present invention;
FIG. 4 is a node merge schematic diagram of an embodiment of the present invention;
FIG. 5 is a schematic view of an encounter with an embodiment of the present invention;
FIG. 6 is a schematic view of an encounter with an embodiment of the present invention;
FIG. 7 is a schematic view of an encounter with an embodiment of the present invention;
FIG. 8 is a schematic view of an encounter with an embodiment of the present invention;
FIG. 9 is a schematic view of an encounter with an embodiment of the present invention;
FIG. 10 is a schematic diagram of a social relationship network of users of an embodiment of the present invention;
FIG. 11 is a schematic diagram of a social relationship network of users with more than 10 encounters and a total duration of more than 10 hours;
FIG. 12 is a schematic diagram of community scale divided by GN algorithm according to an embodiment of the present invention;
fig. 13 is a schematic diagram of community scale divided by FN algorithm according to the embodiment of the invention;
FIG. 14 is a community-scale schematic diagram of CPM algorithm partitioning according to an embodiment of the present invention;
FIG. 15 is a comparative schematic diagram of modularity of GN, FN, CPM algorithm community partitioning in accordance with an embodiment of the present invention;
FIG. 16 is a schematic diagram of a comparison of the run times of three algorithms of an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the following examples in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
As shown in fig. 1, a social relationship modeling method based on a movement track includes the following steps:
1) Recording the track points of the user according to the collected track data of the user in a set time period at preset time intervals to form a time stamp sequence of the track of the user;
the trajectory data of the user is expressed as (< l) 1 ,t 1 ,a 1 ＞,…,＜l i ,t i ,a i ＞,…,＜l n ,t n ,a n >) wherein 1.ltoreq.i.ltoreq.n, l i ＝＜x i ,y i Longitude and latitude data representing the location of the user, t i For user arrival l i Time, a i The motion attribute of the corresponding moment of the track data comprises a direction and a speed; and satisfy t i ＜t i+1 ；
The preset time interval may be an average time interval or a random number selected in a preset time, for example, the preset time interval may be set to be an average time interval of 5s or may be randomly selected in a preset set {2,3,4,5 }.
2) Judging whether the users meet according to the timestamp sequence of each user track, if yes, recording meeting record vector e between every two users ij ＝(v j ,d ij ,n ij ,t ij ) The vector records the user v i And user v j Wherein v is j For user v i Is t ij For user v i And user v j N, n ij For user v i And user v j D is the number of times of meeting ij For user v i And user v j Is the meeting date of (2);
assuming n nodes, when calculating the meeting condition between any two nodes, generating an meeting record vector for each node, wherein each component in the vector is a node v i Encounter information with other n-1 nodes, the structural definition of the record is shown in table 1:
table 1 meet record information
The user v i And user v j Decision-making method for meetingThe formula is as follows:
for user v i And user v j Intercepting the coincident part of the time stamp sequence within the same starting time and ending time range, and the user v i At t 1 The position of the time is point b (la 1 ,lo 1 ) (la is longitude and lo is latitude), at t 2 Is the position of point c (la 3 ,lo 3 ) User v j The position at time t is point a (la 2 ,lo 2 ) And satisfy t 1 ＜t＜t 2 Calculating the Euclidean distance L between a and b and between a and c 1 ，L 2 The method comprises the steps of carrying out a first treatment on the surface of the Setting the distance threshold value as theta D If L 1 ≤θ D And L is 2 ≤θ D Simultaneously calculating the time length delta t of meeting of two users, wherein delta t=t 2 -t 1 Setting the minimum meeting time threshold value as theta T If Deltat is greater than or equal to theta T User v i And user v j If not, the two parts are not met; as shown in fig. 2;
within the same time period, whether the time stamp sequences of the track data overlap or not is a precondition for judging whether two mobile users meet or not, and different overlapping situations are shown in fig. 3.
If the track time stamp sequences of the nodes A and B overlap in a certain time period, a is set as 1 、a 2 For node A at
Position of time, b 1 、b 2 For node B->
The position of the moment, the original track data is preprocessed in the overlapping time period, if +.>
And->
As shown in FIG. 4, then a 1 、a 2 And combining into a time point.
After preprocessing the trajectory data in the whole time period, the time stamp sequence of the node A, B is traversed according to the time sequence, and the following five cases exist.
Case 1:
and->
If->
Calculating the Euclidean distance L 1 ＝D(a 1 ,b 1 )，L 2 ＝D(a 1 ,b 2 ) Length of meeting +.>
As shown in fig. 5.
Case 2:
and->
I.e. the track recording times of the users are completely coincident, as shown in fig. 6. Calculate L 1 ＝D(a 1 ,b 1 )，L 2 ＝D(a 2 ,b 2 )，/>
Case 3:
if->
Calculate L 1 ＝D(a 1 ,b 1 )，L 2 ＝D(a 2 ,b 1 )，
As shown in fig. 7.
Case 4:
if->
Calculate L 1 ＝D(a 2 ,b 1 )，L 2 ＝D(a 2 ,b 2 )，
As shown in fig. 8.
Case 5:
if->
Calculate L 1 ＝D(a 1 ,b 2 )，L 2 ＝D(a 2 ,b 2 )，
As shown in fig. 9.
Each user is regarded as a node, the track data is traversed, and the meeting record vector is obtained;
(1) For each node v i Any other node v= { V is read j } (j=1, 2, …, n, and j+.i) for the same period t s ～t e (t s To start time, t e Is the end time) and converted into a sequence of time stamps s i Where i=1, 2, …, n;
(2) If the time stamp sequence s of two pieces of track data i ，s j K subsequences are overlapped, and for each subsequence, the sequence goes to (3), if all subsequences are processed, goes to (5), otherwise, goes to (2);
(3) Judging the coincidence type and calculating L 1 、L 2 If L 1 ≤θ D And L is 2 ≤θ D Turning to%4) Otherwise, turning to (2);
(4) Calculating delta t, if delta t is more than or equal to theta T If the first time of meeting is the meeting, a node meeting record e is added in the meeting record list ij Otherwise calculate the encounter time t ij ＝t ij +Δt, number of encounters n ij ＝n ij +1; if Deltat < theta T Then go to (2);
(5) Is n node traversals over? If so, go to (6), otherwise go to (1).
(6) Outputting n meeting record vectors E= [ E 1 ,e 2 ,…,e n ]Wherein
If i=j, e ij ＝0。
4) Constructing a social relation model based on the moving track according to the meeting record vector;
from the encounter record vector e= [ E 1 ,e 2 ,…,e n ]The meeting matrix of the user nodes is obtained as follows:
based on the meeting matrix, n is counted ij Component sum t ij The components, calculate the meeting times and the total meeting time of each user node and all other user nodes, get the meeting times matrix k [ i ] of the nodes][j]And an encounter time matrix ti][j]；
The conversion of the meeting times matrix k [ i ] [ j ] into the meeting times probability matrix is as follows:
the probability matrix of the number of encounters between the mobile nodes is as follows:
according to the meeting time t [ i ] [ j ], the meeting time probability matrix among the mobile nodes is calculated as follows:
obtaining the meeting time probability matrix among the mobile nodes according to the above method:
establishing a social relation matrix P SR The following are provided:
P SR ＝(1-ξ)P T +ξ*P N (6)
wherein, xi is an adjusting factor which is more than or equal to 0 and less than 1 and is used for adjusting the meeting frequency matrix P N And an encounter time matrix P T Influence on the social relationship strength measurement between the mobile nodes;
5) And calculating the social relationship strength between the users according to the social relationship model, and dividing the mobile communities to which the users belong.
One example is:
1. experimental data
The gelife project for this experiment contained a GPS track dataset of 182 nodes over five years (4 months 2007 to 8 months 2012). The GPS track of the dataset is represented by a sequence of time stamped recorded points, each point in time containing latitude, longitude and altitude information. The entire dataset contained 17621 tracks, with a total distance of 1292.51 km and a total duration of 50176 hours. These tracks are produced by different GPS recorders and GPS telephones and have multiple sampling rates. 91.5% of the tracks are recorded in a dense representation, such as a track recorded every 1-5 seconds or every 5-10 meters of points, forming a sequence of time stamps.
2 results and analysis
2.1 Community partitioning results based on track data
The social relationship network constructed based on the trajectory data of 182 users is shown in fig. 10, the network belongs to a sparse network, the connection between the nodes is not dense, and isolated nodes are more.
The node number of the network can also be changed by changing the input parameters, for example, if the social relationship network of the user who wants to set the meeting times to be greater than or equal to 10 times and the total meeting time length to be longer than 10 hours, the operation result is shown in fig. 11.
2.2 analysis and comparison of results
Three community discovery algorithms, GN, FN and CPM, are respectively adopted to carry out community division, and the community division results and examples are shown in Table 2.
TABLE 2 Community partition results
Through the GN algorithm, 18 communities were obtained, the number and size of community divisions being shown in fig. 12.
Through the FN algorithm, 20 communities are obtained, and the number and size of community divisions are shown in fig. 13.
The community number obtained by the CPM algorithm is 2. Since CPM is a complete subgraph-based community discovery algorithm, the size of the complete subgraph, i.e., the k value, needs to be known. Typically k is between 4 and 6. Here, the k value is 4 because the experimental network is a sparse small network. The comparison of the community size and the number is shown in fig. 14.
The modularity is a main index for measuring the accuracy of a community discovery algorithm, and the greater the modularity is, the better the community division effect is, and the value range is generally [ -0.5, 1). The modularity of the different algorithms on a given trajectory dataset is shown in fig. 15.
As can be seen from fig. 15, the community division result of the GN algorithm has a higher modularity than that of the FN algorithm, and the size distribution of communities is uneven. The community division result of the FN algorithm has lower modularity, but the community size is uniformly distributed. The modularity of the CPM algorithm is-0.00821, and the modularity is the lowest.
A comparison of the run times of the three algorithms is shown in fig. 16.
From the graph, the CPM algorithm operates at the fastest speed, the GN algorithm operates at the slowest speed, and the FN algorithm operates at the second highest speed. The GN algorithm has higher processing efficiency than the FN algorithm because the GN algorithm continuously partitions the network by calculating the edge betweenness, and the defects of large calculated amount and high time complexity of the GN algorithm are covered when a small network with hundreds of nodes is processed, and the GN algorithm runs faster than the FN algorithm. The FN algorithm is a node merging-based aggregation algorithm, and due to the characteristic, the division result of the algorithm shows the condition that the community size distribution is uniform. In the experiment, because the processed network has smaller scale, the FN algorithm has lower modularity and slower speed than the GN algorithm, and the FN algorithm is more suitable for a large-scale network.
It will be understood that modifications and variations will be apparent to those skilled in the art from the foregoing description, and it is intended that all such modifications and variations be included within the scope of the following claims.
Claims (5)
1. The social relation modeling method based on the movement track is characterized by comprising the following steps of:
1) Recording the track points of the user according to the collected track data of the user in a set time period at preset time intervals to form a time stamp sequence of the track of the user;
the trajectory data of the user is expressed as: (< l) 1 ,t 1 ,a 1 ＞,…,＜l i ,t i ,a i ＞,…,＜l n ,t n ,a n >) wherein 1.ltoreq.i.ltoreq.n, l i ＝＜x i ,y i Longitude and latitude data representing the location of the user, t i For user arrival l i Time, a i Is the motion attribute of the corresponding moment of the track data and satisfies t i ＜t i+1 ；
2) Judging whether the users meet according to the timestamp sequence of each user track, if yes, recording meeting record vector e between every two users ij ＝(v j ,d ij ,n ij ,t ij ) The method comprisesVector records user v i And user v j Wherein v is j For user v i Is t ij For user v i And user v j N, n ij For user v i And user v j D is the number of times of meeting ij For user v i And user v j Is the meeting date of (2);
3) Constructing a social relation model based on the moving track according to the meeting record vector; from the encounter record vector e= [ E 11 ,e 12 ,…,e nn ]The meeting matrix of the user nodes is obtained as follows:
based on the meeting matrix, n is counted ij Component sum t ij The components, calculate the meeting times and the total meeting time of each user node and all other user nodes, get the meeting times matrix k [ i ] of the nodes][j]And an encounter time matrix ti][j]；
The conversion of the meeting times matrix k [ i ] [ j ] into the meeting times probability matrix is as follows:
the probability matrix of the number of encounters between the mobile nodes is as follows:
according to the meeting time t [ i ] [ j ], the meeting time probability matrix among the mobile nodes is calculated as follows:
obtaining the meeting time probability matrix among the mobile nodes according to the above method:
establishing a social relation matrix P SR The following are provided:
P SR ＝(1-ξ)P T +ξ*P N (6)
wherein, xi is an adjusting factor which is more than or equal to 0 and less than 1 and is used for adjusting the probability matrix P of the meeting times N And an encounter time probability matrix P T Influence on the social relationship strength measurement between the mobile nodes;
4) And calculating the social relationship strength between the users according to the social relationship model, and dividing the mobile communities to which the users belong.
2. The method for modeling social relationship based on movement track according to claim 1, wherein in the step 2), the user v i And user v j The manner of judging the meeting is as follows:
for user v i And user v j Intercepting the coincident part of the time stamp sequence within the same starting time and ending time range, and the user v i At t 1 The position of the time is point b (la 1 ,lo 1 ) Where la represents longitude, and lo represents latitude; at t 2 Is the position of point c (la 3 ,lo 3 ) User v j The position at time t is point a (la 2 ,lo 2 ) And satisfy t 1 ＜t＜t 2 Calculating the Euclidean distance L between a and b and between a and c 1 ，L 2 The method comprises the steps of carrying out a first treatment on the surface of the Setting the distance threshold value as theta D If L 1 ≤θ D And L is 2 ≤θ D Simultaneously calculating the time length delta t of meeting of two users, wherein delta t=t 2 -t 1 Setting the minimum meeting time threshold value as theta T If Deltat is greater than or equal to theta T User v i And user v j And if not, the two parts meet.
3. The movement trajectory-based system of claim 1The social relation modeling method is characterized in that step 2) judges whether users meet according to the time stamp sequence of the track of each user, and firstly, preprocessing the time stamp sequence of the track of each user, wherein the preprocessing is as follows: if the track time stamp sequences of the users A and B overlap in a certain time period, a is set 1 、a 2 For user A at
Position of time, b 1 、b 2 For user B +.>
The position of the moment, the original track data is preprocessed in the overlapping time period, if +.>
And->
Will a 1 、a 2 Merging into one time point; after processing the entire time period, both sequences of the user A, B are traversed simultaneously in time-stamped order.
4. The method for modeling social relationships based on movement trajectories according to claim 1, wherein the step 4) uses a community discovery algorithm to divide the mobile communities to which the users belong.
5. The method for modeling social relationship based on movement track according to claim 1, wherein in the step 4), GN algorithm is adopted to divide the mobile communities to which the users belong.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910625340.6A CN110378002B (en) | 2019-07-11 | 2019-07-11 | Social relationship modeling method based on movement track |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910625340.6A CN110378002B (en) | 2019-07-11 | 2019-07-11 | Social relationship modeling method based on movement track |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110378002A CN110378002A (en) | 2019-10-25 |
| CN110378002B true CN110378002B (en) | 2023-05-12 |
Family
ID=68252699
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910625340.6A Active CN110378002B (en) | 2019-07-11 | 2019-07-11 | Social relationship modeling method based on movement track |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110378002B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111026816B (en) * | 2019-12-09 | 2022-09-02 | 重庆锐云科技有限公司 | High-net-value customer group identification method and device based on knowledge graph and storage medium |
| CN110990722B (en) * | 2019-12-19 | 2020-11-06 | 南京柏跃软件有限公司 | Fuzzy co-site analysis method and system based on big data mining |
| CN111460246B (en) * | 2019-12-19 | 2020-12-08 | 南京柏跃软件有限公司 | Real-time activity abnormal person discovery method based on data mining and density detection |
| CN111460332B (en) * | 2020-04-26 | 2020-10-30 | 成都信息工程大学 | Social relationship evaluation method based on user life track |
| CN113873441B (en) * | 2020-06-30 | 2024-02-09 | 香港科技大学 | Method for tracking contact person, and corresponding processing method, system, equipment and medium |
| CN112905907B (en) * | 2021-01-29 | 2021-09-28 | 重庆理工大学 | Dynamic community discovery method for system evolution transplanting partition time sequence network |
| CN112907385A (en) * | 2021-02-23 | 2021-06-04 | 深圳市广信安科技股份有限公司 | Social relationship analysis method, terminal device, and storage medium |
| CN113035366B (en) * | 2021-03-24 | 2023-01-13 | 南方科技大学 | Close contact person identification method, close contact person identification device, electronic device and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012004425A1 (en) * | 2010-07-08 | 2012-01-12 | Telefonica, S.A. | Method for detecting communities in massive social networks using an agglomerative approach |
| CN103853739A (en) * | 2012-11-29 | 2014-06-11 | 中国移动通信集团公司 | Dynamic social relation network community evolution identification and stable community extracting method |
| WO2016188380A1 (en) * | 2015-05-28 | 2016-12-01 | 中兴通讯股份有限公司 | Determination method and apparatus for user equipment |
| CN107071844A (en) * | 2017-06-13 | 2017-08-18 | 湘潭大学 | A kind of opportunistic network routing method divided based on spectral clustering community |
| CN109874159A (en) * | 2019-03-28 | 2019-06-11 | 中南大学 | Moving machine based on comentropy can network node social relationships measurement, cluster foundation and update and method for routing |
-
2019
- 2019-07-11 CN CN201910625340.6A patent/CN110378002B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012004425A1 (en) * | 2010-07-08 | 2012-01-12 | Telefonica, S.A. | Method for detecting communities in massive social networks using an agglomerative approach |
| CN103853739A (en) * | 2012-11-29 | 2014-06-11 | 中国移动通信集团公司 | Dynamic social relation network community evolution identification and stable community extracting method |
| WO2016188380A1 (en) * | 2015-05-28 | 2016-12-01 | 中兴通讯股份有限公司 | Determination method and apparatus for user equipment |
| CN107071844A (en) * | 2017-06-13 | 2017-08-18 | 湘潭大学 | A kind of opportunistic network routing method divided based on spectral clustering community |
| CN109874159A (en) * | 2019-03-28 | 2019-06-11 | 中南大学 | Moving machine based on comentropy can network node social relationships measurement, cluster foundation and update and method for routing |
Non-Patent Citations (1)
| Title |
|---|
| 物联网移动感知中的社会关系认知模型;安健等;《计算机学报》;20120615(第06期);第1164-1174页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110378002A (en) | 2019-10-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110378002B (en) | Social relationship modeling method based on movement track | |
| Chen et al. | Nlpmm: A next location predictor with markov modeling | |
| Wang et al. | When will you arrive? estimating travel time based on deep neural networks | |
| Fan et al. | Citymomentum: an online approach for crowd behavior prediction at a citywide level | |
| Widhalm et al. | Discovering urban activity patterns in cell phone data | |
| CN109035761B (en) | Travel time estimation method based on auxiliary supervised learning | |
| US10831827B2 (en) | Automatic extraction of user mobility behaviors and interaction preferences using spatio-temporal data | |
| Zheng et al. | Diagnosing New York city's noises with ubiquitous data | |
| CN110414732B (en) | Travel future trajectory prediction method and device, storage medium and electronic equipment | |
| Gidófalvi et al. | When and where next: Individual mobility prediction | |
| CN105825297A (en) | Markov-model-based position prediction method | |
| US20160019465A1 (en) | Analyzing Mobile-Device Location Histories To Characterize Consumer Behavior | |
| Chen et al. | Predicting next locations with object clustering and trajectory clustering | |
| CN111444243A (en) | User behavior prediction image method and system based on track information | |
| Wang et al. | Predictability and prediction of human mobility based on application-collected location data | |
| Zhang et al. | Online change-point detection in sparse time series with application to online advertising | |
| CN110162997B (en) | Anonymous privacy protection method based on interpolation points | |
| Zhu et al. | A driving cycle detection approach using map service API | |
| CN111291216B (en) | Method and system for analyzing foothold based on face structured data | |
| Li et al. | A multi-categorical probabilistic approach for short-term bike sharing usage prediction | |
| Dyrmishi et al. | Mobile positioning and trajectory reconstruction based on mobile phone network data: A tentative using particle filter | |
| Anno et al. | Supervised-cityprophet: Towards accurate anomalous crowd prediction | |
| Terroso-Sáenz et al. | Human mobility prediction with region-based flows and water consumption | |
| Liu et al. | An unsupervised collaborative approach to identifying home and work locations | |
| Kang et al. | Traffic anomaly detection via conditional normalizing flow |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |