University libraries are one of the important places to cultivate talents and conduct scientific research, but with the invasion of big data on the Internet, traditional library services cannot accurately understand readers’ needs, leading to a decline in library attendance. To solve this problem, the study proposes to combine multi-view K-mean clustering algorithm with reader behavior analysis to build a college library user portrait system to serve readers. When the enhanced K-mean clustering algorithm from the research was put to the test, the results showed that it performed better than the other two comparison algorithms, with accuracy and loss values of 97 % and 4.3 %, respectively. The user profile method that was suggested in the study was then empirically examined. The findings revealed that, when utilised with university students, the system was effective in raising the attendance rate of students by up to 75 %. In conclusion, it is clear that the method suggested in the study may accurately depict user profiles and offer readers good services, increasing the likelihood that readers will visit university libraries and lowering the waste of educational resources.
原文链接
分享
引用
