Development of Supervised Link Mining Using Computer Vision and Pattern Mining Approach
Seema Mishra, G. C. Nandi · 2016
The network type we have discussed and explored so far in social network analysis is generated from the nonverbal communication like email logs or chats through electronic medium that provide a rich set of data for analysis of social interactions between people. For getting data for these kind of interactions, we could only rely on that available data that sometimes conduces biased information and we ended up with nonrealistic consequences. Therefore, we propose a new direction to be looked into to explore the interactions in network generated from a means other than electronic communication. This mode of interactions is expected to yield results that must be more realistic. In this paper, we address the important problem of discovering and analysis of social networks with the help of computer vision. A new link mining technique is proposed with the help of computer vision and pattern mining approach which exhibits two processes: extraction of social networks and establishing a link between frequently interacted people captured from the video data.