Abstract: On the basis of expounding the concept of big data in science of science, this paper discusses the five opportunities of big data research in science of science: the increasing availability of digital data and new tools, the revelation of the process and laws of scientific discovery, the analysis of scientific structure, relationship and evolution characteristics, the analysis of scientific activities, scientists' behavior and career, and the provision of scientific tools for scientific and technological evaluation and technology prediction research. Five challenges faced by big data in science of science analysis are put forward: the analysis method of scientific structure based on big data in science of science is still immature, new methods, new theories and new algorithms of scientific measurement and evaluation based on big data in science of science need to be developed urgently, the big data in science of science analysis platform and tools are relatively few, the technical means of integrating multiple data are not mature, and the universality of science big data revealing law still needs to be tested.

Key words: Big data in science of science, Science of science, Scientometrics, Science and technology evaluation, Scientific structure