Improving the Performance of Video Content Genuineness Using Convolution Neural Network

Video searching in search engines uses metadata information to find the relevant videos according to the search queries. Metadata information mainly comprises the title and description of the video. The major drawback of this approach is that it overlooks whether or not the content of the video is genuine or not. Since the metadata information is provided by the uploader, the person may provide false information about it. Therefore, there is a need of improving the results of video searched. The proposed work classifies the video in different categories and then compares the tag provided to each video with the tags that were extracted from the metadata of the video. The other factor like views count, likes and dislikes, comments is also considered for the ranking of the video searched. It improves the genuineness of the content of the video searched.