Abstract: This paper is to explore discovery service users' information behavior by employing web analytics and big data approach. Taking Xi'an Jiaotong-Liverpool University as an example, massive data on how users access to discovery service and interact with the system have been captured in order to objectively generate multi-dimensional information and identify users' general characteristics of information behavior. The finding reveals user information preferences which can be provided references for library resource integration and discovery tool optimization.

Key words: web analytics approach; Google analytics; discovery service; users information behavior