Abstract. Academic libraries have offered ebooks for some time, however little is known about how readers interact with them while making relevance decisions. In this paper we seek to address that gap by analyzing ebook transaction logs for books in a university library.Keywords: ebooks, log analysis, book selection, HCI, information behavior
IntroductionConsider the process of borrowing a book from a library (digital or physical): the reader searches or browses the collection to identify candidate books that are potentially relevant; they each candidate book to assess its actual relevance; and they borrow those books deemed relevant to explore them more depth at a later date. Of course, finding a useful book is rarely so straightforward: these stages may be sequential or interleaved. In this paper, we consider a specific aspect of this process: the examination of a candidate book after the candidate has been identified in the library collection; we focus specifically on ebooks. How a reader explores physical books when making relevance decisions-which parts of books are viewed, how quickly, in what orderhas been relatively neglected [8], likely because such a study in a physical domain would be intrusive and 'creepy' [16]. To our knowledge this aspect of ebook selection has also not been explored, surprisingly, given that the availability of ebook transaction logs allows such a study to be conducted post hoc so as to avoid disturbing the experience of those using an ebook collection. This paper describes exactly such a study based on the transaction logs of a university library ebook collection.In Section 2 of this paper we explore the previous research on ebook usage and book selection (both physical and electronic); in Section 3 we describe the ebook collection on which our study is based and gives details of the log sample we analyzed. Section 4 presents the results of our analysis, including which parts of the