User Profile Based Paper Recommendation System

Abstract

With the spread of science and the increasing number of researchers working in academic fields, there has also been a significant increase in the number of academic publications. Researchers always follow new works published for keeping their knowledge up to date. However, due to thousands of academic publications published every day from many academic sources, academics are not always able to find publications about their subjects. Today, almost all of online academic databases employ a recommendation module that only considers the studies similar to the paper that the user looked at. However, a recommendation system based on the information of a single article is often not enough. In this study, the proposed method recommends by considering user's publications, user's co-authors and co-authors' papers. Therefore, metadata of the articles published by the researcher in the past are evaluated as time-awareness by the method we proposed. In this way, the most relevant articles to the user's profile can be found by using the proposed method in the data repository created from the exact contents of hundreds of thousands of academic works. The method uses TF-IDF frequency-based similarity analysis method. In the evaluation phase, the performance of the proposed method was examined. The accuracy of the method was measured by several different tests. The results are very promising and demonstrate that the method can produce accurate and quality results.

Influence of tweets and diversification on serendipitous research paper recommender systems

“…Other researches have employed stereotyping, item-centric recommendations, and hybrid recommendations. In this article, we employ a content-based approach, as a number of works have done in the past with promising results (Sugiyama & Kan, 2010;Nascimento et al, 2011;Achakulvisut et al, 2016;Kaya, 2018).…”

“…To help researchers overcome the problem of information overload, various studies have developed recommender systems (Beel et al, 2016;Bai et al, 2019). Recommendations are generated based on considerations such as a user's own papers (Sugiyama & Kan, 2010;Kaya, 2018) or the papers a user has accessed or liked in the past (Nascimento et al, 2011;Achakulvisut et al, 2016). Most previous studies have focused only on improving the accuracy of recommendations, one example of which is normalised discounted cumulative gain (nDCG).…”

Peer Review #1 of "Influence of tweets and diversification on serendipitous research paper recommender systems (v0.1)"

“…Other researches have employed stereotyping, item-centric recommendations, and hybrid recommendations. In this article, we employ a content-based approach, as a number of works have done in the past with promising results (Sugiyama and Kan, 2010;Nascimento et al, 2011;Achakulvisut et al, 2016;Kaya, 2018).…”

“…To help researchers overcome the problem of information overload, various studies have developed recommender systems (Beel et al, 2016;Bai et al, 2019). Recommendations are generated based on considerations such as a user's own papers (Sugiyama and Kan, 2010;Kaya, 2018) or the papers a user has accessed or liked in the past (Nascimento et al, 2011;Achakulvisut et al, 2016). Most previous studies have focused only on improving the accuracy of recommendations, one example of which is normalised discounted cumulative gain (nDCG).…”

A Collaborative Approach Toward Scientific Paper Recommendation Using Citation Context

“…Different contents have been used in the literature. Some of them are, authors [3] extract title and abstract from a single paper, authors [4] extract whole contents of a list of papers authored by an author, authors [5] extract papers' keywords and so on to provide recommendation. Most of these approaches extract various contents from the scientific papers to build user profile to provide recommendations.…”