Application recommending method and apparatus

  • US 10,679,132 B2
  • Filed: 03/14/2018
  • Issued: 06/09/2020
  • Est. Priority Date: 03/28/2014
  • Status: Active Grant
First Claim
Patent Images

1. An application recommending method, comprising:

  • acquiring a first behavior eigenvalue according to behavior data of a first user, wherein the first behavior eigenvalue represents the behavior data of the first user in a numerical form;

    determining at least one second user from a user relationship chain of the first user according to the first behavior eigenvalue, the second user being on the user relationship chain of the first user, a degree of similarity between a second behavior eigenvalue and the first behavior eigenvalue being greater than a preset threshold, and the second behavior eigenvalue being a behavior eigenvalue of the second user;

    determining a to-be-recommended application for the first user on a basis of behavior data of the at least one second user; and

    acquiring a scheduling weight of each compute node before computing a system load of the compute node,wherein the scheduling weight of the compute node is determined in accordance with;

    wi=(1/Zi)/(1/Z1+1/Z2+ . . . +1/Zn),wherein wi indicates a scheduling weight of the ith compute node, Z1 indicates a system load of a first compute node, Z2 indicates a system load of a second compute node, and Zn indicates a system load of a nth compute node.

View all claims

    Thank you for your feedback