Topic-based awareness computing model for video-sharing service

A video-sharing service, where users can upload videos to, have rapidly spread recently. The number of videos in the service has increased dramatically as well as the number of users. With the rapid growth of the number of videos, we are faced with is having to arrange and categorize videos. However, it is difficult to let a computer aware the contents (or topics) of videos. This paper focused on such video-sharing services and proposed a novel framework of a system which improves the awareness of the outline of video retrieval result. Our proposed system detects topics from a set of retrieval matched videos utilizing time data, analyzes topic properties using author's diversity and offers an interface which lets person to grasp the outline of video retrieval result in one glance. We also presented an experimental result of applying our proposed methods for some sets of video retrieval result.