Market basket analysis of library circulation data

Abstract

"Market basket analysis" algorithms have recently seen widespread use in analyzing consumer purchasing patterns - specifically, in detecting products that are frequently purchased. We apply the a-priori market basket tool to the task of detecting subject classification categories that co-occur in transaction records of books borrowed from a university library. This information can be useful in directing users to additional portions of the collection that may contain documents which are relevant to their information need, and in determining a library's physical layout. These results can also provide insight into the degree of "scatter" that the classification scheme induces in a particular collection of documents.