Semantic Web link analysis to discover social relationships in academic communities

To discover how precisely links among researchers' Web sites capture the ties of academic communities, this study conducts Web citation analysis. 3,878 Web sites of computer scientists and their 8,811 articles are assessed. 200 most frequently cited researchers from them are subjected to Web author cocitation analysis and compared to bibliographical data. As a result, a cluster of researchers on the Web becomes a superset of bibliographical one, that is, a community representing a larger research field. We then develop Web link ontology that defines various relationships among researchers to apply metadata annotation to the links among 200 sites. The findings are 1) since intellectual ties analysis separates off sociocognitive ties, it results into a better approximation of bibliographical author cocitation, 2) since sociocognitive ties analysis can utilize various links not only coauthor networks, diverse kinds of social connections can be found, 3) merging intellectual ties and sociocognitive ties analyses detect gatekeepers who bridges research fields on social networks.