Aiming at the problem of single angle of description and unintuitive presentation of information for university scientific researchers, we integrate multi-source heterogeneous data, such as information of scientific researchers, projects, papers, monographs, awards, patents and standards to draw user portraits of scientific researchers in university and accurately portray the "thousands of people and thousands of faces", comprehensively understand the information of scientific researchers, and objectively evaluate their research level, in order to improve the efficiency and level of scientific research management, and provide references for precise scientific research services and scientific decision-making.