No open-access location is known for this record yet.
No open-access location is known for this record yet.
Scientific journals can capture a scholar’s research career. A researcher’s publication data often reflects his/her research interests and their social relations. It is demonstrated that scientist collaboration networks can be constructed based on co-authorship data from journal papers. The problem with such a network is that researchers are limited within their professional social network. This work proposes the idea of constructing a researcher’s social network based on data harvested from metadata of scientific publications and personal online profiles. We hypothesize that data, such as, publication keywords, personal interests, the themes of the conferences where papers are published, and co-authors of the papers, either directly or indirectly represent the authors’ research interests, and by measuring the similarity between these data we are able to construct a researcher social network. Based on the four types of data mentioned above, social network graphs were plotted, studied and analyzed. These graphs were then evaluated by the researchers themselves by giving ratings. Based on this evaluation, we estimated the weight for each type of data, in order to blend all data together to construct one ideal researcher’s social network. Interestingly, our results showed that a graph based on publication’s keywords were more representative than the one based on publication’s co-authorship. The findings from the evaluation were used to propose a dynamic social network data model.
Loading…
Imported from OpenAlex (CC0) — nothing here is generated.