Methods and systems for user behavior based feature extraction and personalized recommendation

Methods and systems for user behavior based feature extraction and personalized recommendation

  • CN 104,239,324 A
  • Filed: 06/17/2013
  • Published: 12/24/2014
  • Est. Priority Date: 06/17/2013
  • Status: Active Application
First Claim
Patent Images

1. based on a feature extracting method for user behavior, it is characterized in that, comprising:

  • Collect the primitive behavior information of user, the primitive behavior information according to described user generates user behavior data point;

    Described user behavior data point comprises the dimension score value of user ID, dimension mark and correspondence;

    Adopt dimension score value to carry out cluster to described user behavior data point according to described dimension mark, form multiple Cluster space;

    Binaryzation is carried out respectively for the dimension score value of user behavior data point in described multiple Cluster space;

    Result extraction according to binaryzation meets the characteristic dimension of one or more dimensions marks as Cluster space of preset requirement.

View all claims
    ×
    ×

    Thank you for your feedback

    ×
    ×