Rank-mediated collaborative tagging recommendation service using video-tag relationship prediction

A great many tags and videos are shared and created by a mass of distributors on Web 2.0 video sharing sites. This increasing user-generated content can further benefit service innovation of collaborative tagging. In order to enhance efficient video retrieval and online video marketing (OVM) application, this research proposes a rank-mediated collaborative tagging recommendation service that allows the distributors predicting the ranks of video retrieval from the shared video archive using vote-promotion algorithm (VPA). The system experiments evaluate the number of tags and videos between simple text retrieval and VPA. The user surveys verify the relevance, helpfulness, and satisfaction of the recommended tags. From the perspectives of service innovation, this research is to develop a systematic and quantified a video-tag relationship prediction and recommendation self-service that can provide an intelligent collaborative tagging service on video sharing sites.