Developing an Integrated Topic Modeling Approach for Text-Based Knowledge Accumulation
Maximilian Weinhold, Dries Faems · Academy of Management Proceedings · 2024
This paper introduces an integrative topic modeling approach, leveraging recent advancements in natural language processing (NLP) for knowledge accumulation. It addresses the limitations of traditional topic modeling methods (i.e., Latent Dirichlet Allocation), employing novel algorithms such as BERTopic, CorEx, and leveraging large language models like GPT. Using 2567 journal paper abstracts in the research domain of innovation ecosystems, the paper identifies and analyzes 27 distinct topics, enabling a comprehensive bibliographic review. This innovative approach showcases the potential of a combination of various topic modeling techniques for scalable, contextual text analysis in management research.