Method and device for determining personal commute time period

A device for determining a personal commute time period, comprising:

• one or more processors;
• and
• a memory, configured to store one or more programs, when executed by the one or more processors, cause the one or more processors to perform a method for determining a personal commute time period, comprising:
• collecting positioning data of a user within a preset time period, the positioning data being Global Positioning System (GPS) positioning data of a mobile terminal that is collected according to a preset frequency and stored when the user's geographical location changes;
• determining a location of home and a location of workplace of the user based on positioning data;
• and
• ranking the positioning data based on time, selecting positioning data corresponding to locations between the location of the home and the location of the workplace, and generating data of the at least one commute trajectory according to the selected out positioning data, the at least one commute trajectory respectively corresponding to at least one day within the preset time period, and each commute trajectory is a trajectory from the home to the work place of the user;
• performing, by the one or more computing devices, cluster analysis on the data of the at least one commute trajectory to obtain at least one commute trajectory cluster, and obtaining a combined trajectory by combining the commute trajectories in the corresponding commute trajectory cluster;
• wherein performing, by the one or more computing devices, cluster analysis on the data of the at least one commute trajectory to obtain at least one commute trajectory cluster comprises:
• S1, initializing a commute trajectory cluster according to data of each commute trajectory, to obtain at least one commute trajectory cluster;
• S2, determining whether a number of the at least one commute trajectory cluster is greater than 1;
• S3, in response to that the number of the at least one commute trajectory cluster is greater than 1, obtaining a similarity between each two commute trajectory clusters;
• S4, determining whether a smallest similarity of the obtained similarities is greater than a preset similarity threshold;
• S5, in response to that the smallest similarity is smaller than or equal to the preset similarity threshold, combining the two commute trajectory clusters that have the smallest similarity into one commute trajectory cluster;
• and
• S6, repeating S2-S5, until the number of the at least one commute trajectory cluster is smaller than or equal to 1 or the smallest similarity is greater than the preset similarity threshold;
• for each commute trajectory T of the at least one commute trajectory, determining, by the one or more computing devices, a commute speed of the user according to a combined trajectory corresponding to a commute trajectory cluster to which the commute trajectory T belongs;
• determining, by the one or more computing devices, a commute time period corresponding to the commute trajectory T according to the commute speed;
• and
• displaying each commute trajectory T and the commute time period corresponding to each commute trajectory T on a display screen of the mobile terminal;
• wherein the at least one commute trajectory corresponds to at least one day respectively, and initializing a commute trajectory cluster according to data of each commute trajectory, to obtain at least one commute trajectory cluster comprises:
• for a commute trajectory Ti of the ith day, ranking data of record points of the commute trajectory Ti based on time, wherein, 1≤i≤N, N is the number of the at least one commute trajectory, data of each record point comprises a user identification uid, a longitude coordinate lng, a latitude coordinate lat, and a timestamp time;
• determining a location array Pi of the commute trajectory Ti based on the longitude coordinates and the latitude coordinates of the record points of the commute trajectory Ti, and determining a time array Di of used time period between two adjacent record points of the commute trajectory Ti;
• and
• generating a commute trajectory cluster Ci based on the commute trajectory Ti, the location array Pi, and the time array Di, where, Ci=<<Pi, Di>, {Ti}>, the location array Pi, and the time array Di forms a combined commute trajectory <Pi, Di> of the commute trajectory cluster Ci;
• wherein the similarity between two commute trajectory clusters is determined according to following formula:
• DM(A,B)=L(M)/max(L(A),L(B))
• where, DM(A, B) represents the similarity between commute trajectories A and B, M represents a combined commute trajectory of commute trajectories A and B, L(M) represents a length of the combined commute trajectory M, L(A) represents a length of the commute trajectory A, L(B) represents a length of the commute trajectory B.