PubMed related articles: a probabilistic topic-based model for content similarity

• Main Authors: Lin Jimmy, Wilbur W John
• Format: Article
• Language: English
• Published: BMC 2007-10-01
• Series: BMC Bioinformatics
• Online Access: http://www.biomedcentral.com/1471-2105/8/423

<p>Abstract</p> <p>Background</p> <p>We present a probabilistic topic-based model for content similarity called <it>pmra </it>that underlies the related article search feature in PubMed. Whether or not a document is about a particular topic is computed from term frequencies, modeled as Poisson distributions. Unlike previous probabilistic retrieval models, we do not attempt to estimate relevance–but rather our focus is "relatedness", the probability that a user would want to examine a particular document given known interest in another. We also describe a novel technique for estimating parameters that does not require human relevance judgments; instead, the process is based on the existence of MeSH ^®in MEDLINE ^®.</p> <p>Results</p> <p>The <it>pmra </it>retrieval model was compared against <it>bm25</it>, a competitive probabilistic model that shares theoretical similarities. Experiments using the test collection from the TREC 2005 genomics track shows a small but statistically significant improvement of <it>pmra </it>over <it>bm25 </it>in terms of precision.</p> <p>Conclusion</p> <p>Our experiments suggest that the <it>pmra </it>model provides an effective ranking algorithm for related article search.</p>