Abstract / Description of output
Libraries are sources of large-scale data: both in terms of their collections and the information they collate on their spaces, users, and systems. These data provide opportunities to explore technical, methodological, and ethical questions from the valuable interdisciplinary perspective of Data Science and the Digital Humanities. In light of this, we will explore our analysis of library datasets using Subjectify 1 , an automatic classification matching tool developed to assist analysis of UK Non-Print Legal Deposit (NPLD) collections. NPLD regulations were introduced to the UK in 2013 to support legal deposit libraries to collect electronic publications (2013). Access restrictions mean that readers may only use these materials on fixed terminals within the physical walls of the six legal deposit libraries (see British Library, 2014 for details). The resultant web logs are therefore unambiguous sources of NPLD collection usage within UK legal deposit libraries.
Our study is part of an established tradition of user studies in the digital humanities. To date, these have focused on user behaviour with digital resources (Warwick et al., 2008; Ross and Terras, 2011; Sinn and Soares, 2014). Web log analysis has been used successfully in this context for over twenty years (Almind and Ingwersen, 1997; Nicholas et al., 2005; Gooding, 2016). These studies adopt methodological approaches and topics of study that contribute directly to our understanding of information sources in the digital humanities. However, there have been fewer studies that address critical humanistic perspectives to inform approaches to the data itself. This paper addresses that gap by describing our research into the users of NPLD materials in the United Kingdom, and the implications of automatic classification matching for library dataset analysis. It will address the following questions: what insights into users of digital library collections can be derived from automatic classification matching? What limitations are introduced by the use of existing classification schemes? And, in light of ongoing debates on responsible data curation in DH (Weingart, 2014; Brown et al., 2016), how might DH and LIS scholars collaborate to inform ethical analysis of large-scale library datasets?
Our study is part of an established tradition of user studies in the digital humanities. To date, these have focused on user behaviour with digital resources (Warwick et al., 2008; Ross and Terras, 2011; Sinn and Soares, 2014). Web log analysis has been used successfully in this context for over twenty years (Almind and Ingwersen, 1997; Nicholas et al., 2005; Gooding, 2016). These studies adopt methodological approaches and topics of study that contribute directly to our understanding of information sources in the digital humanities. However, there have been fewer studies that address critical humanistic perspectives to inform approaches to the data itself. This paper addresses that gap by describing our research into the users of NPLD materials in the United Kingdom, and the implications of automatic classification matching for library dataset analysis. It will address the following questions: what insights into users of digital library collections can be derived from automatic classification matching? What limitations are introduced by the use of existing classification schemes? And, in light of ongoing debates on responsible data curation in DH (Weingart, 2014; Brown et al., 2016), how might DH and LIS scholars collaborate to inform ethical analysis of large-scale library datasets?
Original language | English |
---|---|
Publication status | Published - 10 Jul 2019 |
Event | Digital Humanities 2019: Annual conference of ADHO (Assoc of Digital Humanities Organisations) - Utrecht University, Utrecht, Netherlands Duration: 9 Jul 2019 → 12 Jul 2019 |
Conference
Conference | Digital Humanities 2019 |
---|---|
Country/Territory | Netherlands |
City | Utrecht |
Period | 9/07/19 → 12/07/19 |