Technical evolution and prediction of blockchain based on different evolution patterns by text mining and bibliometric methods

Given the disruptive changes that blockchain has brought to various industries, the profound evolutionary paths of blockchain technology (BCT) can help to comprehend it better. Thus, this paper investigates BCT from a technology management perspective that includes three parts: 1) topic identification; 2) topic evolution analysis; 3) topic prediction. In the first part, the major application fields and annual topics of BCT are identified based on cluster analysis and the hierarchical Dirichlet process (HDP), respectively. The second part calculates the topic importance and similarities amongst annual topics. Five evolutionary patterns were obtained based on the similarity relations. In the last part, both emerging and lasting topics are predicted by considering their importance and evolutionary patterns. The findings are substantial and interesting, including five main BCT-application fields, several emerging and lasting technical topics and the two longest evolutionary paths.