Research on the Identification and Evolution of Patent Thicket based on Semantic Enhancement

This paper introduces a semantic enhanced patent thicket recognition and evolution method, and identifies the advantages of different patent owners through patent owner classification analysis. It uses the triangle citation method to detect patent thickets, and uses TF-IDF and K-means to reveal mainstream technology trends and patent owner dominance through semantic clustering analysis of patent abstracts. It also combines lifecycle theory to track the evolution of patent thickets. The effectiveness of the method has been demonstrated through empirical research in the VR field, and the evolution process of the patent thicket in the VR field and the sustained advantages of patent holders have been identified.