Doc-to-Doc Recommender for Medical Literature with Similarity of Molecule Graphs

The increasing literature leads to formidable pressure for medical researchers. Most existing recommender approaches mainly depend on text-based information. How to extract and utilize the heterogeneous information, especially the graphic ones, to improve the recommender is worthy of further exploring. To this end, we establish a document-to-document recommender system for medical literature (D2D-MR). Specifically, we proposed HB-GED, the Half-branch GED algorithm, and the bipartite-graph-based algorithm for solving the molecule similarity and the paper similarity, respectively. Experimental results on real-world datasets demonstrate the effectiveness of the proposed recommender system.