Evaluating Asynchronous Sharing of Links and Annotation Sessions as Social Interactions on Internet Videos

Social video sites have become popular on the Internet, and enable users to post, share, annotate, and comment online videos. Ubiquitous computing applications may be exploited in this scenario to allow users to experience novel forms of social interactions. The literature presents guidelines and metrics for the evaluation of the sociability in online video applications, but the evaluation of the underlying social interactions have not been reported. In this paper, social interactions are specified as behavioral contingencies - modeled as if-then rules which describe what people do - and measured using data mining procedures. We present a novel technique for codifying and evaluating social interactions involving the asynchronous sharing of links and annotations on Internet videos. We report results from applying our technique in a user study.