Time Sensitive Ranking with Application to Publication Search

Link-based ranking has contributed significantly to the success of Web search. PageRank and HITS are the best known link-based ranking algorithms. These algorithms do not consider an important dimension, the temporal dimension. They favor older pages because these pages have many in-links accumulated over time. Bringing new and quality pages to the users is important because most users want the latest information. Existing remedies to PageRank are mostly heuristic approaches. This paper investigates the temporal aspect of ranking with application to publication search, and proposes a principled method based on the stationary probability distribution of the Markov chain. The proposed techniques are evaluated empirically using a large collection of high energy particle physics publication. The results show that the proposed methods are highly effective.