Altmetrics for Digital Libraries : Concepts, Applications, Evaluation, and Recommendations
The volume of scientific literature is rapidly increasing, which has led to researchers becoming overloaded by the number of articles that they have available for reading and difficulties in estimating their quality and relevance (e.g., based on their research interests). Library portals, in these circumstances, are increasingly getting more relevant by using quality indicators that can help researchers during their research discovery process. Several evaluation methods (e.g., citations, Journal Impact Factor, and peer-reviews) have been used and suggested by library portals to help researchers filter out the relevant articles (e.g., articles that have received high citations) for their needs. However, in some cases, these methods have been criticized, and a number of weaknesses have been identified and discussed. For example, citations usually take a long time to appear, and some articles that are important can remain uncited. With the growing presence of social media today, new alternative indicators, known as “altmetrics,” have been encountered and proposed as complementary indicators to traditional measures (i.e., bibliometrics). They can help to identify the online attention received by articles, which might act as a further indicator for research assessment. One often mentioned advantage of these alternative indicators is, for example, that they appear much faster compared to citations. A large number of studies have explored altmetrics for different disciplines, but few studies have reported about altmetrics in the fields of Economics and Business Studies. Furthermore, no studies can be found so far that analyzed altmetrics within these disciplines with respect to libraries and information overload. Thus, this thesis explores opportunities for introducing altmetrics as new method for filtering relevant articles (in library portals) within the discipline of Economic and Business Studies literature. To achieve this objective, we have worked on four main aspects of investigating altmetrics and altmetrics data, respectively, of which the results can be used to fill the gap in this field of research. (1) We first highlight to what extent altmetric information from the two altmetric providers Mendeley and Altmetric.com is present within the journals of Economics and Business Studies. Based on the coverage, we demonstrate that altmetrics data are sparse in these disciplines, and when considering altmetrics data for real-world applications (e.g., in libraries), higher aggregation levels, such as journal level, can overcome their sparsity well. (2) We perform and discuss the correlations of citations on article and journal levels between different types and sources of altmetrics. We could show that Mendeley counts are positive and strongly correlated with citation counts on both article and journal levels, whereas other indicators such as Twitter counts and Altmetric Attention Score are significantly correlated only on journal level. With these correlations, we could suggest Mendeley counts for Economic and Business Studies journals/articles as an alternative indicator to citations. (3) In conjunction with the findings related to altmetrics in Economics and Business Studies journals, we discuss three use cases derived from three ZBW personas in terms of altmetrics. We investigate the use of altmetrics data for potential users with interests in new trends, social media platforms and journal rankings. (4) We investigated the behavior of economic researchers using a survey by exploring the usefulness of different altmetrics on journal level while they make decisions for selecting one article for reading. According to the user evaluation results, we demonstrate that altmetrics are not well known and understood by the economic community. However, this does not mean that these indicators are not helpful at all to economists. Instead, it brings forward the problem of how to introduce altmetrics to the economic community in the right way using which characteristics (e.g., as visible numbers attached at library records or behind the library’s relevance ranking system). Considering the aforementioned findings of this thesis, we can suggest several forms of presenting altmetric information in library portals, using EconBiz as the proof-of-concept, with the intention to assist both researchers and libraries to identify relevant journals or articles (e.g., highly mentioned online and recently published) for their need and to cope with the information overload.