Based on the SERVQUAL evaluation model, this paper examines the characteristics and service model of university archive services and establishes a knowledge service model framework for university smart archives. The service framework utilizes information fusion technology to quantitatively analyze the archive situation through the combination of 6 dimensions and 23 secondary indicators. One of the key points is to combine the whitening weight function of upper and lower limit measurements and the central whitening weight function to form a grey clustering algorithm in order to solve the “cliff” effect of the whitening weight function of the traditional upper and lower limit measurements. After the development of the system was completed, the index weights were first calculated by means of a questionnaire and then practiced in the school. The results show that responsiveness has the highest weight, 0.8612; security is second, 0.6321. And tangibility is the last, only 0.1792. The first college scored 0.461 in C1, and the second college scored only 1.87 in E1, which pulled the average score of the first-level index to 4.52. After the practice, the survey proved that the system meets the needs of the college and can point out areas for archive improvement. Archives improvement direction.