Paper Info
Title
Technology Forecasting Based on Semantic and Citation Analysis of Patents: A Case of Robotics Domain
Abstract
In modern society, science and technology has become the core competitiveness of a company or a country. Discovering the development trajectory a technological domain is of great significance. In this article, we propose a framework to answer this question by analyzing the semantics of patent texts and the citation relationships among patents, because patents in a specific technological domain provide rich resources of technological development process. First, we collect the patent data of a specific domain from the database of United States Patent and Trademark Office. Second, we extract different topics from these patents’ texts by Latent Dirichlet Allocation. Third, an index for evaluating the correlation between two patents is defined according to the semantic similarities and citation relationship between them. Fourth, we found the global and local important patents through the global and local important index (LII). Finally, we discover some community of the network constructed by the important patents filtered by LII and statistical analysis methods to obtain some meaningful results. Furthermore, we take the patents in the technological domain of robotics as an example to examine the proposed method and reveal the technological development trends in this domain.
Year
DOI
Venue
2022
10.1109/TEM.2020.2978849
IEEE Transactions on Engineering Management
Keywords
DocType
Volume
Main important path analysis,patent citation,robotics,text semantic similarity
Journal
69
Issue
ISSN
Citations 
4
0018-9391
0
PageRank 
References 
Authors
0.34
17
2
Name
Order
Citations
PageRank
Zhipeng Qiu100.34
Zheng Wang26412.16