Multidimensional Scientometric indicators for the detection of emerging research topics

The detection of emerging research topics (ERTs) is an important step in the promotion of potentially promising research. In this paper, we first distinguish the concept of ERT from that of common-related topics. We discuss multi-dimensional and practical bibliometric indicators to discover high-impact ERTs at a fine level of granularity. To this end, we focus on three objectives. First, we determine a method for identifying research topics from scholarly publications. Second, we developed a method for uncovering ERTs. Third, we conduct an experimental analysis to demonstrate the operationalization of our method. High-impact ERTs are selected based on the advice of field experts. Then, we divide the identified ERTs into seven patterns and propose different research and development (R&D) strategies for each topic. In general, previous studies have mainly focused on the attributes of “novelty” and “growth.” Compared with extant ERT identification analysis, this study pays more attention to the future economic and social impact of ERTs and the reduction of uncertainty. This article is the first attempt to devise an ERT detection framework by comprehensively measuring its five characteristics (radical novelty, relatively fast growth, persistence and coherence, potential high impact, and uncertainty and ambiguity reduction) and proposing feasible operational indicators and processes. Thus, different R&D layout strategies for different economic and social-impact topics are outlined herein, which contribute significantly to the development of science and technology policies. The evidence presented suggests that our methodology can be used by decision makers to evaluate emerging scientific research.