Diverse Relevance Ranking in Web Scrapping for Multimedia Answering

With exponential growth of media padding on the web, the skill to hunt for media entities not just based on text footnote, but also visual content has become important. The main problem is the lack of diversity of generated media data. To enhance the above said process, a scheme is proposed to enrich text answers with image and video and able to take relevance and diversity into account by exploring the content of images. Our Scheme investigates a rich set of techniques: question/answer categorization, query creation, image and video hunt re-ranking. This approach automatically determines which type of media information should be added for textual answers with high precision and to make enrich media data more diverse.