MedRank: Discovering Influential Medical Treatments from Literature by Information Network Analysis

Chen, L., Li, X. and Han, J.

    Medical literature has been an important information source for clinical professionals. As the body of medical literature expands rapidly, keeping this knowledge up-to-date becomes a challenge for medical professionals. One question is that for a given disease how can we find the most influential treatments currently available from online medical publications? In this paper we propose MedRank, a new network-based algorithm that ranks heterogeneous objects in a medical information network. The network is extracted from MEDLINE, a large collection of semi-structured medical literature. Different types of objects such as journal articles, pathological symptoms, diseases, clinical trials, treatments, authors, and journals are linked together through their relationships. The experimental results are compared with the expert rankings collected from doctors and two baseline methods, namely degree centrality and NetClus. The evaluation shows that our algorithm is effective and efficient. The success of categorized entity ranking in medical literature domain suggests a new methodology and a potential success in ranking semi-structured data in other domains.
Cite as: Chen, L., Li, X. and Han, J. (2013). MedRank: Discovering Influential Medical Treatments from Literature by Information Network Analysis. In Proc. Database Technologies 2013 (ADC 2013) Adelaide, Australia. CRPIT, 137. Wang, H. and Zhang, R. Eds., ACS. 3-13
pdf (from crpit.com) pdf (local if available) BibTeX EndNote GS