Academic Knowledge Sharing in Research Social Networks


Student thesis: Doctoral Thesis

View graph of relations


Related Research Unit(s)


Awarding Institution
  • Jian MA (Supervisor)
  • Zhongsheng Hua (External person) (External Supervisor)
Award date6 Jan 2020


Academic information overload has presented a challenge to researchers given the rapid growth of scientific articles. Methods have been proposed to help professional readers find relevant articles on the basis of their publications. Although effectively sharing publications is essential to spreading knowledge and ideas, scant studies have focused on knowledge sharing from an author perspective. In reality, authors can only share their articles through a few offline channels. A large number of articles can be accessed through search engines, but they are not necessarily shared with readers after publication. This scenario leads to the unfortunate phenomenon of many undiscovered papers. The rise of research social networks has provided researchers with a new and effective platform for knowledge sharing and discovery under open access policy. Authors can upload and share their publications with their peers and friends on-line. Few research has studied how to effectively share articles on professional social networks from an author perspective.

This study leverages online research social network to propose a general academic knowledge sharing framework to assist authors in effectively sharing their publications with interested readers. Two academic knowledge sharing strategies are developed to meet different demands. For authors who want to share their articles with situations that are particularly similar to their own research, a similarity-based academic knowledge sharing framework is proposed. For authors who want to share their articles with situations that are relevant to their own research but not particularly similar, a relevance-based academic knowledge sharing framework is put forward. In our approach, researcher-level and document-level analyses are integrated in the same model, which works in two stages. 1) Researcher-level analysis combines research topic, social relation and research quality dimensions. 2) Document-level analysis includes short-term research interests extracted from online social behaviour and long-term research interests drawn from researchers’ publications. Our research is developed and verified in ScholarMate, a prevalent research social network platform. Compared with other baseline methods (TF-LDF, LDA, LDA-traSR and RLA), our approach significantly improves the accuracy and satisfaction of recommendation results. Our method can also disseminate papers efficiently to readers who have no publications.

    Research areas

  • academic knowledge sharing, recommender systems, research social network