期刊名称:International Journal of Computer Science & Technology
印刷版ISSN:2229-4333
电子版ISSN:0976-8491
出版年度:2012
卷号:3
期号:4
页码:494-496
语种:English
出版社:Ayushmaan Technologies
摘要:The technologies are growing continually and due to this there are a large network of social networking and media collection library is increased. Particularly searching for a video from media libraries on the network has become a challenging task. Presently video search is mainly based on text, titles, descriptions and image features associated with it. There are many methods have been developed to improve the video search performance but they don’t provide high accuracy on top ranked documents. In this paper we present a novel framework that integrates multiple features and help us to improve the video search performance in terms of relatedness of documents. We use semantic mapping and feedback policy to gain high accuracy on top ranked result. The proposed framework may be the most promising framework to gain high accuracy on the top ranked documents. We will provide the result on the basis of two performance parameters namely lost query result and ghost query result.