Font Size: a A A

Research Of Patent Tendency Analysis Based On Entity Relationship

Posted on:2015-11-22Degree:MasterType:Thesis
Country:ChinaCandidate:Y F DuFull Text:PDF
GTID:2298330467967056Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
The patent analysis method includes the quantitative analysis method and the qualitativemethod. The traditional patent analysis is based on quantitative analysis method, however, itcannot represent how technologies are used in their technology area, and it can’t indicate howthey affect other technologies. This paper adopts the qualitative analysis method based onentity relationship to make up the shortcomings of the quantitative analysis.Based on entity relationship, this paper explores the patent analysis, includes thefollowing aspects:Explore and research the patent similarity. This paper suggests a method for measuringpatents’ similarity based on Subject-Action-Object (SAO). Based on the OLLIE, the latestentity information extraction tool, the rule process is added so that the better SAO structure isextracted. Because a large number of items exist and the verb phrases are simple and clear inpatent, a method for measuring patents’similarity based on SAO is presented. And at last, thispaper merges the VSM module into SAO module and gets a good result.Explore and research the patent tendency analysis. In this paper, the patents in the fieldof mechanical engineering are researched, so the patent analysis is mainly concentrated onproduct components. This paper works at the clustering of patent, the reorganization ofpatent’s product, etc. based on entity relationship. This paper researches the tendency analysisbased on component.Based on the research of structure of patent texts and SAO structure, this paper shows apatent tendency analysis system. By the experiment, this system can get the better results thanother systems.
Keywords/Search Tags:PatentAnalysis, Entity Relationship, Patent Similarity, Tendency Analysis
PDF Full Text Request
Related items