Technology Convergence Assessment by an Integrated Approach of BERT Topic Modeling and Association Rule Mining
Priyanka Chand Bhatt, Yu-Chun Hsu, Kuei-Kuei Lai, Vinayak A. Drave · IEEE Transactions on Engineering Management · 2025
The rapid evolution of technology necessitates advanced methods to assess and understand emerging innovations. As formal documents record inventions, patents provide rich data for analyzing technological advancements. This study employs text mining and data mining techniques to analyze patent data, focusing on technology convergence and innovation trends in e-payment technological domain. Using BERT (Bidirectional Encoder Representations from Transformers) topic modeling, patent abstracts are classified into distinct thematic areas, uncovering hidden patterns and thematic landscapes of technological domains. International Patent Classification codes categorize these patents, facilitating the identification of technological convergence through Association Rule Mining. The study integrates these methods, addressing gaps in previous research by providing a comprehensive analysis of technological evolution and convergence. The research aims to propose a Convergence Indicator, to highlight heterogeneous technological convergence. The limitations of study rely on patent data, suggesting future research incorporate additional data sources for a more holistic view of technological convergence. The findings underscore the potential of integrating text mining and data mining techniques in technology assessment, contributing to the understanding of technological evolution and convergence dynamics.