Technology scanning: a systematic mapping review using text analytics

Abstract

Abstract In rapidly changing technological environments, companies must effectively identify and evaluate technological changes to sustain their dynamic capabilities. The growing body of research on technology scanning and innovation trajectories makes it difficult to perform systematic literature reviews using traditional methods alone. Many existing reviews rely on citation-based indicators or manual screening procedures, which can limit scalability and make it difficult to capture the evolving structure of the literature. This paper presents a text-analytics systematic mapping review of the technology-scanning literature. The workflow combines corpus construction, machine-learning–based screening, topic modeling, clustering, and similarity-based mapping to support the systematic analysis of research on technology scanning and technological change detection. Using a large collection of publications retrieved from Web of Science and Scopus, the workflow presented in this paper identifies methodological approaches, thematic structure, and emerging analytical trends in the technology scanning literature, including human-in-the-loop topic modeling, semi-supervised and guided topic modeling, and reinforcement learning-based approaches. The resulting topic clusters and their associated documents are analyzed and mapped to three main research questions (A, B, and C) and eleven corresponding sub-questions presented in Table 2. The results show that existing studies are rich in concepts but fragmented, with most models relying on predefined variables and lacking adaptive mechanisms to capture technological change. Several approaches are also semi-automated and rely on human interpretation. Even recent AI-driven approaches that generate analytical outputs are not connected to transparent analytical workflows and do not support continuous knowledge updates. The proposed workflow demonstrates how topic modeling and similarity-based analysis can be incorporated into a systematic mapping review of technology-scanning research. The findings highlight opportunities for developing more adaptive and continuously updated approaches to technology scanning.

The authors' abstract, as published at the source. Scientometrics, 2026 · DOI ↗