bX Scholarly Article Recommender Service: Overview


Introduction


Welcome to bX, the Ex Libris scholarly article recommender service, which will provide your institution's researchers with recommendations about scholarly articles-recommendations based on the individual scholar's area of interest. An ever-expanding corpus of link-resolver activities carried out by scholars from prestigious research institutions around the world lies at the hub of the bX service. These recommendations are created through the structural analysis of logs generated from SFX linking activity by users at research institutions around the world. Tens of millions of such logs have already been collected in the bX database.

You are probably familiar with similar services on commercial sites such as Amazon. On Amazon, you see recommendations for books; bX provides recommendations for scholarly journal articles. For example, a user who searched for the article "Equity in the finance of health care: Some international comparisons" received the following recommendations from the bX service:
bX Recommendation

While many searches for scholarly articles yield recommendations, other searches do not, but as the bX database grows, so, too, will the number of articles for which recommendations can be offered.

The bX recommender service has evolved from well-documented work carried out through the collaboration of the Ex Libris bX team and researchers from the Los Alamos National Laboratory, Johan Bollen and Herbert Van de Sompel (see J. Bollen and H. Van de Sompel, "An architecture for the aggregation and analysis of scholarly usage data," 24 May 2006; available at http://public.lanl.gov/herbertv/papers/jcdl06_accepted_version.pdf