Incorporating circulation data in relevancy rankings for search algorithms in library collections

Abstract

This paper demonstrates a series of analyses to calculate new clusters of shared subject headings among items in a library collection. The paper establishes a method of reconstituting anonymous circulation data from a library catalog into separate user transactions. The transaction data is incorporated into subject analyses that use supercomputing resources to generate predictive network analyses and visualizations of subject areas searched by library users. The paper develops several methods for ranking these subject headings, and shows how the analyses will be extended on supercomputing resources for information retrieval research.