Georgia Kapitsaki¹ and Dragan Ivanovic², ¹University of Cyprus, Cyprus and ²University of Novi Sad, Serbia

Many systems provide search and recommendation capabilities to scholars that search for scientific documents including research papers and dissertations. The appearance of search results may largely affect the system use. Traditional approaches provide textual formats for showing the results to users, whereas more recent approaches concentrate on other forms, e.g., on two dimensions. Moreover, this presentation may be adapted to user needs providing a personalised user experience combined with other contextual factors, such as enriching user search with keywords from recently used documents. In this paper, we present our work on results representation in the framework of a dissertation search engine in the Serbian language with the ultimate aim to provide a more personalised experience to users. We have integrated our approach in the PhD UNS digital library system of the University of Novi Sad, a research information, library and educational information system, and are discussing as early evaluation how users are perceiving this approach outlining also our vision for a context-aware digital library system. The initial results demonstrate the usefulness of providing more choices to the users adapting application to their needs.

PHD UNS, word cloud, search results, log analysis