Data analytics and search behavior

Dr. Niu will present two recent research projects: Exploring shoppers’ search behavior on a large e-commerce website As e-commerce is ubiquitously available these days, customers could shop online anytime anywhere using any smart devices. Shoppers’ searching and browsing behavior and the conversion related to these behaviors are rarely studied. In this study, we have teamed up with a large e-commerce website and analyzed one-day search logs in 2014. Different measures were compared among three different kinds of platforms—desktops, tablets and mobile phones. The discrepancies as well as similarities regarding search time, search location, click behaviors and query behaviors are reported. The implications for better online shopping experiences are discussed at the end.   Leveraging text analytics to implement serendipity in health information recommendation As web-based health information consumption increases dramatically due to its broad availability and accessibility, we aims at identifying novel ways to support more effective health information browsing by supporting “serendipitous” discoveries that can match unmet information needs of web users. To this end, the we developed association mining and text modeling techniques that will enable the current and next generation of web applications to deliver richer and “surprising” information that users did not look for in the first place, but that are very relevant to their unexpressed needs.   Refreshments will be served.