Page 135 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 43

ZHAO Xing / Exploring the measurement features of usage data for academic literature 135

academia and industrial circles also constitutes a part of literature influence. For an academic
publication without a high citation count, as long as it has a high usage count, its actual influence
would still be worth noting. For discipline difference, the literature of some applied engineering,
humanities and social science disciplines may have a relatively low citation density, but those
achievements that have an actual guiding effect for industrial circles or enjoy an extensive
influence can still be identified from usage data. As for behavioral motive, although usage data
cannot be expected to fully make up for the shortage of citation data, the usage data can still
provide different perspectives, that is, the features of usage behavior can serve as a supplement to
citation behavior.
On that account, the mining of the data on academic literature usage has drawn close attention of
bibliometric circles. In this regard, the inclusion of download data into research and applications
provides a very important signal. As early as 2005, Moed (2005) and Bollen, Van, Smith, and
Luce (2005) had already begun to pay attention to the relationship between download count and
citation count for journals and literature. China’s CNKI database has also become one of the
pioneering full-text databases in releasing the download counts of electronic literature (Wan,
Hua, Rousseau, & Sun, 2010). However, since then, the progress in the research of usage data
and the application development of evaluation have been slow, mainly because of the absence of
data published on internationally recognized and relatively authoritative platforms, such as the
citation of the WoS platform and the impact factor of Journal Citation Report (JCR). As a result,
the empirical studies by scholars on usage data were constantly limited to samples constituted
by a few samples until very recently (Schloegl & Gorraiz, 2010; Jamali & Nikzad, 2011;Schlögl,
Gorraiz, Gumpenberger, Jack, & Kraker, 2014). After 2010, Priem, Taraborelli, Groth, & Neylon
(2010) put forward Altmetrics, which exerted a very profound influence on the bibliometrics circle
(You & Tang, 2013; Qiu &Yu, 2015). Notwithstanding the limitations to or disputes over its logical
framework (Bornmann, 2014; Moore, 2016), the assiduous explorations by Altmetrics of academic
influence on a wider scale and from more dimensions have clarified a very important objective of
future development for bibliometrics. Usage data also fall within the scope of the investigation of
Altmetrics. However, it should be noted that research on usage data as an academic achievement
carrier have a very long history and can be traced back to the usage analysis on the library users
of paper literature on Library Science. It is also worth noting that the analysis environment also
differs from the social media emphasized by Altmetrics, as has been incisively described by
Glanzel and Gorraiz (2015). To sum up, neither the preliminary research on usage data nor the
emergence and promotion of Altmetrics can solve the problem of unifying the data source of
academic literature usage. This problem has become a critical bottleneck restraining the large-scale
inclusion of usage data into research and applications.
2) Potential values and theoretical features of the usage data of the WoS platform
With this background, the usage data given by the WoS platform is of important potential value.
Their possible influence on the development and evaluation application of bibliometrics mainly