Large-scale news topic tracking and key-scene ranking with video near-duplicate constraints
Xiaomeng Wu, Ichiro Ide, Shin’ichi Satoh · 2009
To make full use of the overwhelming volume of news videos available today, it is necessary to track the development of news stories from different channels, mine their dependencies, and organize them in a semantic way. We propose a novel news topic tracking and re-ranking system. The main contributions include: (1) a novel scheme of mining topic-related stories through tracking and re-ranking on the basis of near duplicates built on top of text, (2) a proposed simple but effective query-expansion algorithm for improving the representativeness of a search query, (3) a large-scale broadcast video database containing more than 34,000 news stories constructed for experimentation, and (4) a novel key-scene ranking scheme for analyzing both text similarity and video near-duplicate constraints.