Abstract: Scientific research performance evaluation is an important means to improve the efficiency of scientific research activities and optimize the allocation of resources. With the rapid growth of performance data and continuous strengthening of information correlation, existing evaluation methods face new challenges in efficiency and associated knowledge discovery. This paper used basic data and objective data to construct the basic knowledge map, and completed the knowledge through extended knowledge to form a multi-source integration knowledge map for scientific research performance evaluation. It was beneficial to carry out the evaluation of the input and output analysis of scientific research performance, evaluate the influencing factors of benefit, the rationality of performance evaluation and the stability of research direction, and had great significance to the scientific research performance evaluation.