×

Method and apparatus for extracting entity names and their relations

  • US 9,430,742 B2
  • Filed: 06/02/2014
  • Issued: 08/30/2016
  • Est. Priority Date: 09/28/2000
  • Status: Expired due to Fees
First Claim
Patent Images

1. A method comprising:

  • generating a number of Information-Gain (IG)-Trees based on a memory learning technique and training sets relating to a document, wherein the training sets are based on global context features, wherein a global context feature offers a broader view of a word or a word sequence with regard to an entirety of the document;

    extracting entity names and relations between entity names based on the IG-Trees;

    receiving annotated data;

    parsing, at least partially, the annotated data, wherein parsing includes identifying syntactic structure of sentences within the annotated data; and

    extracting the training sets from the parsed annotated data, wherein the training sets are further based on features including one or more of local context features, surface linguistic features, and deep linguistic features, wherein the global context feature, when included in an IG-Tree, further offers a set of first verbs in a same sentence that appear before or after the word or the word sequence for the entirety of the document or corpus.

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