An Academic Search Engine for Personalized Rankings

Rapidly increasing information on the Internet and the World Wide Web can lead to information overload. Search engines become important tools to help WWW users to discover information. Exponential increases in published research papers, academic search engines become indispensable tools to search for papers in their expertise and related fields. In order to improve the quality of search, an academic search engines’ capability should be enhanced. This paper proposes a search engine for personalized rankings. In order to evaluate the performance of personalized rankings, thirty-five graduate students from the Department of Web Engineering and Mobile Application Development at Dhurakij Pundit University are participants in the research experiment. Participants are asked to use a prototype of an academic search engine to find and bookmark any research papers according to their interests, which would guarantee that each participants’ list of interesting research papers could be recorded. Normalized Discounted Cumulative Gain (NDCG) is used as a metric to determine the performance of the personalized rankings. The experiments suggest that the personalized rankings outperform the original search rankings. Hence, the proposed academic search engine with personalized ranking benefits research paper discovery.