Abstract
We are able to analyze the meaning in the news video by using various data, such as a moving picture, an audio, a caption, and so on. In order to browse the news video effectively, classification and skimming of the news articles are very essential. In this paper, we propose both the automatic classification and skimming methods of the news articles using the closed-caption. The automatic classification method uses tags in the closed-caption for the purpose of distinction of speakers. The skimming method extracts the part of the article introduced by an anchor in the closed-caption as a representative sentence, and also extracts the representative frames consisted of the anchor frame, open-caption frames, and frames synchronized with the frequently appeared terms.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
