Abstract: Convergence research is a new research paradigm based on interdisciplinary approaches to solving major economic and social problems. In recent years, it has become a development trend that major countries pay great attention to. This paper first analyzes and reveals the characteristics of convergence research from the perspectives of ontology, epistemology and methodology, which illustrate the difference between the convergence research paradigm and the traditional research paradigm. Then this paper puts forward an entropy index, i.e., convergence entropy, to measure convergence research, and takes brain science research as an example to carry out empirical research with this index. There are mainly four theoretical results and conclusions from this paper. Firstly, convergence research is especially a holistic reflection on the ontology, epistemology and methodology of scientific research, instead of the paradigm shifting on the level of specific disciplines. Secondly, from the ontological point of view, the convergence research activities are mainly about natural ontology or social ontology, which means that the research problems or objects of convergence research are generally derived from the major challenges of nature and economic society, rather than simply deduced from the existing knowledge system. Thirdly, from the perspective of epistemology, convergence research is a typical constructivism, which not only requires knowledge construction at the individual level, but also emphasizes proactive, systematic and coordinated management and application of knowledge at the organizational and national levels. Fourthly, in terms of methodology, convergence research emphasizes wider openness and deeper integration. On the one hand, convergence research absorbs different disciplines according to the specific research problems, instead of the interaction of the same certain disciplines, which means that different research problems are solved by the convergence of different disciplines. In other words, convergence research involves a larger variety of disciplines at a deeper extent of interactions among these disciplines. On the other hand, convergence research will form a new research framework and research language through the "chemistry" interaction between disciplines, and eventually develop into convergence fields, rather than forming a new branch of disciplines. This paper also has important empirical results, which are mainly about the index of measuring convergence research. In detail, by using the index of convergence entropy to examine the development of brain science research from 1945 to 2020, it can be seen that global brain science research has experienced at least four development stages featured by their convergence entropy values during the 76 years. From 1945 to 1969 is the first development stage, showing a significant upward trend of convergence entropy. From 1970 to 1990 is the second development stage which means that global brain science research entered the plain period of multi-disciplinary integration, showing a relatively steady value of convergence entropy. From 1991 to 2011 is the third development stage, with the convergence entropy value of brain science research entering the second period of rapid increase, indicating that the multi-disciplinary convergence phenomenon in brain science research is more active than before. Since 2012, global brain science research has been in the fourth development stage, with the convergence entropy has been increasing rapidly again, after the brief plain period, which reflects that convergence research not only includes more and more disciplines in quantity, but also has a deeper degree of interdisciplinary integration. The above empirical results are highly consistent with the results of expert qualitative judgments made in 2011 in the USA, which concluded the three major changes in the field of life science research over the last 60 years. This empirical analysis indicates that the index of "convergence entropy" is scientific and feasible to quantitatively identify the convergence research and describe its convergence research stages.Finally, this study helps to promote the formation of consensus on the basic characteristics of convergence research in the science community, and removes cognitive barriers for policy making related to convergence research. It also lays a methodological foundation for the situation analysis and planning of convergence research. Of course, empirical studies on the index of "convergence entropy" in this paper are only conducted on the example of brain science. Whether the index needs to be revised when applied to other convergence fields needs to be furtherly studied, which is also important to promote the further development of convergence research theories and methods.

Key words: convergence research, characteristics, measurement, convergence entropy