Patent key component extraction with the application of patent similarity analysis

被引：0

作者：

Hu, Po ^{[1
]}

Huang, Minlie ^{[1
]}

Zhu, Xiaoyan ^{[1
]}

机构：

[1] Department of Computer Science and Technology, Tsinghua University, Beijing

来源：

Journal of Computational Information Systems | 2014年 / 10卷 / 13期

关键词：

Dependency tree; Natural language processing; Patent key component; Patent similarity;

D O I：

10.12733/jcis11279

中图分类号：

学科分类号：

摘要：

Patents are crucial for a company to protect its intellectual properties. Effective text mining in patent portfolios provides companies with valuable insights to develop marketing strategies. In this paper, we study the problem of patent key component extraction, and demonstrate its application value by patent similarity analysis. We propose a syntactic-based approach to identify the key components in an invention, and integrate them with topic models to measure patent similarity. Experimental results on real-world patent portfolios show the effectiveness of the proposed method. © 2014 by Binary Information Press

引用

页码：5813 / 5820

页数：7