Related paper recommendation to support online-browsing of research papers
Manabu Ohta, Toshihiro Hachiki, Atsuhiro Takasu · 2011
An online-browsing support system for research papers has been developed that extracts technical terms from a paper and presents links to useful pages such as those explaining the terms. A method to further use the extracted technical terms is proposed to recommend papers to a user that are related to the paper he or she is browsing. Specifically, the proposed method generates a bipartite graph consisting of papers retrieved by the extracted technical terms, which are called related papers, and technical terms appearing in these related papers. It then ranks the related papers using the HITS algorithm for analyzing the bipartite graph and recommends top-ranked papers to the user. The proposed method was compared with other recommendation methods in terms of effectiveness in an experiment.