×

Complex service network ranking and clustering

  • US 10,210,558 B2
  • Filed: 02/17/2017
  • Issued: 02/19/2019
  • Est. Priority Date: 06/26/2014
  • Status: Active Grant
First Claim
Patent Images

1. A method comprising:

  • carrying out offline functionality-based co-ranking and clustering on a knowledge base, said knowledge base characterizing a heterogeneous information technology services network comprising a plurality of services, a plurality of providers, and a plurality of attributes;

    storing results of said functionality-based co-ranking and clustering as annotations of said services and said providers in said knowledge base, to obtain an annotated knowledge base;

    obtaining a service requirement from a customer requiring information technology services;

    querying said annotated knowledge base, based on said service requirement; and

    returning to said customer an ordered list of at least given ones of said services, based on said querying,wherein;

    said clustering comprises service influence based probabilistic clustering using a unified neighborhood random walk distance measure, wherein service influence accounts for social interactions between object vertices and social activities between object vertices and associated attribute vertices, wherein object vertices represent services or providers and attribute vertices denote their properties; and

    said co-ranking comprises service influence based ranking.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×