Lend me some sugar: Borrowing rates of neighbouring books as evidence for browsing

Abstract

There is more to choosing a book than simply keyword searching. Browsing is a fundamental part of the information seeking process, and one that information seekers profess to value, though it has attracted little study. This dearth of research is undoubtedly in part because browsing is nebulous and difficult to quantify. In this paper we use a large circulation dataset from an academic library consortium to examine whether books in the library stacks are loaned in clusters, with a view firstly to confirming the existence of book browsing that has been reported anecdotally, and secondly to quantifying its impact on loan patterns.

Books' Interest Grading and Fiction Readers' Search Actions During Query Reformulation Intervals

“…Thus, providing a diverse range of metadata elements in title pages supports readers to judge the books. In addition, options for efficient browsing could enhance the finding of similarly interesting titles [16]. In our study identifying interesting titles by pivot browsing was often experienced as frustrating by the participants.…”

Tyranny of Distance: Understanding Academic Library Browsing by Refining the Neighbour Effect

“…This paper extends the methodology of our previous work [9]. We will briefly review our dataset, then describe four new tests employed to examine the patterns of co-borrowing.…”

“…The only recent study of the impact of browsing-or more specifically shelf location-on loans within academic libraries is our own [9]. We used a large publicly available circulation dataset from the OCLC [34], and selected six libraries based on a set of criteria to ensure their broad similarity.…”

“…Not every visit to the library will result in co-borrowings, and borrowing rates per book are low (annual circulation ranged from 0.08 to 0.18 loans per book for all collections [9]). The data shows a log-log distribution often seen in information behaviour [37].…”

“…Our prior work [9] leveraged two established digital library techniques-examining physical libraries [1,6,7] and transaction log analysis [10,11]-to determine whether physical layout of library shelves affects book borrowing patterns. In that work we specifically examined whether proximity to a loaned book increases the chance a book will itself be loaned: we found a strongly significant increase in this likelihood that we termed the neighbour effect.…”