×

Clustering method for a point of interest and related apparatus

  • US 10,423,728 B2
  • Filed: 05/06/2016
  • Issued: 09/24/2019
  • Est. Priority Date: 11/07/2013
  • Status: Active Grant
First Claim
Patent Images

1. A clustering method, comprising:

  • acquiring, from a location tracking device, a locating point set of a user within a preset period, wherein each locating point in the locating point set indicates a particular location of the user at a particular time;

    generating a stay point set according to the locating point set, wherein each stay point in the stay point set represents a hot area, and the hot area meets a set of conditions, the set of conditions comprising;

    a distance between geographic locations of any two locating points in the hot area is less than a higher locating precision in locating precisions of the two locating points, and a maximum value of a time interval between locating points in the hot area is greater than a preset time threshold;

    acquiring movement states of locating points comprised in a hot area represented by each stay point;

    calculating a confidence level of each stay point in the stay point set according to the movement states of the locating points comprised in the hot area represented by each stay point and according to a relation between confidence level weights of movement states of each stay point in the stay point set and a quantity of locating points that are located in the hot area represented by a corresponding stay point and whose movement state correspond to a corresponding movement state, wherein each movement state corresponds to a confidence level weight, wherein a lower average speed corresponding to the movement states of the locating points in a hot area represented by a stay point indicates a higher confidence level of the stay point, and wherein a lower movement speed corresponding to a movement state indicates a larger confidence level weight of the movement state;

    obtaining a trusted stay point from the stay point set by screening according to the confidence level of each stay point in the stay point set, wherein a confidence level of the trusted stay point is greater than a preset confidence level threshold; and

    clustering density-connected trusted stay points to form a point of interest, wherein the density-connected trusted stay points are trusted stay points that represent hot areas whose ranges are connected to each other.

View all claims
  • 2 Assignments
Timeline View
Assignment View
    ×
    ×