Identifying Emerging Trends of Blockchain Technology Using a Topic-based Patent Mining Model

Abstract

In this research, we apply a latent Dirichlet allocation (LDA) model with noun phrase extraction to identify the underlying structure of collected patent corpus related to Blockchain technology. In order to uncover the emerging topics, we go one step further by utilizing the document-topic matrix generated from LDA to weigh each topic popularity and growth rate. Also, integrating the unstructured data such as applicants, forward citations, claims to illustrate the competitive landscape, patent value and technology evolution map. Our results reveal the future trends in the Blockchain technology field and find major firms have been focused on different technology topics. Finally, through the patent map combining the analyzed outcomes can help Blockchain companies formulate the patent strategy and facilitate decision process.