Incorporating Full Text and Bibliographic Features to Improve Scholarly Journal Recommendation
Tirthankar Ghosal, Ananya Chakraborty, Ravi Sonam, Asif Ekbal, Sriparna Saha, Pushpak Bhattacharyya · 2019
Selecting an appropriate venue to communicate one's research is the very first step in scholarly communication. Many papers are simply rejected from the editor's desk on not being submitted to the right journal. Existing journal recommender systems extract keywords only from the title and abstract sections of candidate articles and produce journal recommendations based on their weighted-match with a domain-specific vocabulary. Here in this work, we investigate a simple yet effective approach by incorporating additional information from bibliography and body section of academic manuscripts and show their potency to yield a better recommendation. On a closed set of ten different journals, our content-based recommender achieves significant improvement over the usual baselines (at least ~ 10%). Our preliminary approach is simple yet promising and if suitably applied could efficiently recommend journals from a larger pool as well.