BUILDING A SEMANTIC RESEARCHGATE CLONE USING OPEN SOURCE TECHNOLOGIES

ResearchGate is a free to join social networking site for scientists and researchers that enables the sharing and querying about publications, connecting with colleagues, seeking for new collaborators, obtaining statistics and metrics about publications, asking questions and getting answers from researchers that share similar interests; it also has job seeking capabilities. It has a high usability factor and incorporates many elements of familiar social media sites such as Facebook, Twitter and LinkedIn. Members can track and follow the research publications of others in their field. ResearchGate claims to have more than 3 million users, although it is not clear how many of these are active accounts that are maintained and updated regularly.

In this paper we describe the design and the development of a publications management system, based exclusively on open source technologies, tailored to specific academic needs, which can be a valuable source of information regarding the research performance of the institute. The system has the same capabilities as ResearchGate, with emphasis on social features and a high usability factor. It obtains data automatically from online sources, like PubMed or Scopus, and uses a linear classifier to associate these data to the user’s profile. End users are able to edit their publications and correct any false identifications with minimum effort. The system produces automatic real time reports that can be used in the internal and external evaluation procedures of the institute. The data are described using The Bibliographic Ontology and exposed as open linked data, through a SPARQL end-point. Code snippets are available to embed publications on any web site, e.g. for use in an e-portfolio. The evaluation of the system shows a high usability factor. Finally, a set of possible future steps aiming at improving the system efficiency are discussed.