With the advents of internet, the importance of electronic resources is growing. Due to the increasing expensiveness of electronic resources, university libraries normally received budgets from parent institutions annually. They necessarily applied effective and systematic methods for decision making in electronic resources purchase or re-subscription. However, there are some difficulties in practices: First of all, libraries are unable to receive user records; second, the COUNTER statistics does not include details about users and their affiliation. As a result, one cannot conduct advanced user analysis based on the usage of users, institutions, and departments. To overcome the difficulties, this study presents a feasible model to analyze electronic resource usage effectively and flexibly. We set up a proxy server to collect actual u sage raw data. B y analyzing items in internet browsing records, associated with original library automatic system, this study aims at exploring how to use effective ways to analyze big data of website log data. We also propose the process of how original data to be transformed, cleared, integrated, and demonstrated. This study adopted a medical university library and its subscription of medical electronic resources as a case. Our data analysis includes (1) year of subscription, (2) title of journal, (3) affiliation, (4) subjects, and (5) specific journal requirements, etc. The findings of the study are contributed to obtain further understanding in policy making and user behavior analysis. The integrated data provides multiple applications in informatics research, information behavior, bibliomining, presenting diverse views and extended issues for further discussion.