中科微点-全球专利检索

  1. Login
  2. Sign Up
Invention Grant
US07769749B2 Web page categorization using graph-based term selection 有权
使用基于图的术语选择的网页分类

  • Patent Title: Web page categorization using graph-based term selection
  • Patent Title (中): 使用基于图的术语选择的网页分类
  • Application No.: US11938983
    Application Date: 2007-11-13
  • Publication No.: US07769749B2
    Publication Date: 2010-08-03
  • Inventor: Xiaofei He
  • Applicant: Xiaofei He
  • Applicant Address: US CA Sunnyvale
  • Assignee: Yahoo! Inc.
  • Current Assignee: Yahoo! Inc.
  • Current Assignee Address: US CA Sunnyvale
  • Agency: Greenberg Traurig, LLP
  • Agent James J. DeCarlo
  • Main IPC: G06F17/30
  • IPC: G06F17/30
Web page categorization using graph-based term selection
Abstract:
This disclosure describes systems and methods for categorizing web pages. Web pages and terms selected from those web pages are organized in a matrix. The number of terms in the matrix are filtered using a Laplacian score algorithm. A linear regression algorithm or some other algorithm may use the filtered set of terms to fit the web pages into pre-defined categories.
Abstract(Chinese):

本公开描述了用于对网页进行分类的系统和方法。 从这些网页中选择的网页和术语被组织成矩阵。 矩阵中的项数使用拉普拉斯算法进行滤波。 线性回归算法或一些其他算法可以使用过滤的术语集来将网页适合于预定义的类别。

Public/Granted literature
Information query
Espacenet
Patent Agency Ranking
0/0