Year: 2009
Type: conference-paper
Abstract: The dynamic nature of citation networks makes the task of ranking scientific articles hard.Citation networks are continually evolving because articles obtain new citations every day.For ranking scientific articles, we can define the popularity or prestige of a paper based on the number of past citations at the user query time; however, we argue that what is most useful is the expected future references.We define a new measure, FutureRank, which is the expected future PageRank score based on citations that will be obtained in the future.In addition to making use of the citation network, FutureRank uses the authorship network and the publication time of the article in order to predict future citations.Our experiments compare FutureRank with existing approaches, and show that FutureRank is accurate and useful for finding and ranking publications.
Authors: Hassan Sayyadi, Lise Getoor
Institution: University of Maryland, College Park
Language: English
Subfield: Statistics, Probability and Uncertainty
Field: Decision Sciences
Domain: Social Sciences
Open Access status: bronze
Funder: National Science Foundation
OpenAlex ID: W95321676