On computing similarity in academic literature data: Methods and evaluation

Abstract

Similarity computation for academic literature data is one of the interesting topics that have been discussed recently in information retrieval and data mining. Consequently, a variety of methods has been proposed to compute the similarity of scientific papers. In this paper, we present various similarity methods and evaluate their effectiveness via extensive experiments on a real-world dataset of scientific papers.