In the practice of early university evaluation, due to the backwardness of information technology and evaluation methods, the following problems were brought about: the social needs were neglected, the evaluation subject was single, the evaluation method was simple, and the evaluation data source was single. In the paper a big data model of university evaluation is presented, which studies the educational activities from 8 dimensions: core performance, innovation ability, teamwork, resources, audience willingness, parental evaluation, employer evaluation and social contribution. The model takes account of long-term, short-term factors, individual and overall factors, internal and external factors, educators and audience factors, etc., and carries out omnidirectional analysis. Every dimension is analyzed from 2 aspects: positive and negative. Then based on the model, an University Evaluation System based on Big Data (UESBD) was designed and implemented. Finally, the software was used to analyze the data of 20 colleges and universities from multiple dimensions to evaluate the educational activities of education.