Identifying technology convergence across FinTech: an integrated approach of text mining and association rule mining

Purpose Financial technology (FinTech) has significantly transformed and advanced the traditional financial services sector. The aim of this study is to identify the technology convergence (TC) and innovation trends utilizing both text and data mining techniques on patent data within the FinTech domain. Design/methodology/approach A total of 36,335 patents in FinTech were retrieved for analysis from the Lens database. Bidirectional Encoder Representations from Transformers topic modelling was performed on the Abstract and Claims Text Dataset. Association rule mining (ARM) was performed on 204 distinct Cooperative PClassification (CPC) codes among the retrieved patent data. The interestingness measures calculated were confidence, Kulczynski and imbalance ratio (IR). Findings Top five topics obtained revealed the convergence of Fintech into blockchain, advertising, information processing, logistics and geolocation-based payment. Findings also highlighted the organizations such as Bank of America, Capital One, IBM, Amazon and Mastercard being the key players in the topics obtained. ARM revealed 405 association rules across 31 distinct CPC codes, revealing convergence across data and image processing, along with machine learning, biometrics and a significant number of patents in energy management systems and renewable energy. Practical implications BERT topic modelling, combined with ARM’s interestingness measures, provides financial organizations with a detailed map of structured insights into the strength and direction of technological intersections. Originality/value The scarcity of research on patents and TC across FinTech innovations provides an opportunity for this research to be a novel approach in identifying heterogeneous technological convergence in the FinTech domain.