Domain Specific Publication Recommendation System using Deep Learning Approach
6 Pages Posted: 29 Aug 2019
Date Written: August 28, 2019
Recommendation system is any system that automatically suggests content or items for users and also helps the users to distinguish particular items that best match their interests or preferences. Computer science and information technology influences our daily life broadly and deeply. In the computer science journals or conferences, a vast amount of papers are being provided day by day. To help authors to choose where they ought to submit their manuscript, this system presents the content-based publication recommender system on the field of computer science. This work recommends suitable journals or conferences based on the abstract of paper prepared by author. The entire work is divided in to two parts: first one is to identify the areas of abstract like Machine learning, Artificial Intelligence, Computation Language, Neural and evolutionary computation, Computer Vision and Pattern Recognition etc., by using multi-label classification with deep learning; second one is to recommend journals or conferences by making rules. The recommended journals or conferences are evaluated by comparing the accuracy of the system with existing work and manual evaluation has also being adopted. The proposed system outperforms the state-of-art technique.
Keywords: Recommender System, Content-based publication recommender system, Multi-label classification
Suggested Citation: Suggested Citation