Mapping Innovation Domains Using Topic Modeling

Abstract

Innovation plays a crucial role in driving economic growth and development, improving competitiveness, and creating new opportunities for individuals and businesses. It can lead to higher living standards and improved quality of life by addressing some of the world's most pressing challenges, such as climate change, health crises, and poverty. In recent years, artificial intelligence (AI) has emerged as one of the most influential forces shaping everyday life. It is also revolutionizing the innovation process by supporting R&D activities, automating complex tasks and enabling data analysis at speeds far exceeding human capabilities. The adoption of AI technologies accelerates the development of new products and business models while fundamentally transforming decision-making processes. Consequently, companies are required to restructure their innovation processes in response to rapid technological advancement and evolving workforce roles. AI is widely perceived as a source of unlimited possibilities, and its increasing adoption is strongly reflected in the expanding body of scholarly work. The number of papers addressing innovation and AI has grown significantly in recent years. To illustrate, assess and map research at the intersection of AI and innovation, this study analyzes published work indexed in the Elsevier Scopus databases. The research examines thousands of publications related to innovation by analyzing titles, abstracts, and keywords, employing BERTopic modeling and bipartite graph analysis to identify emerging patterns. The findings reveal five distinct innovation technology domains and systematically map their relationships with various AI methods. The results indicate that AI-related innovations are not isolated phenomena but are widely integrated across multiple domains, highlighting the importance of crossdepartmental collaboration to maximize system-wide benefits.