An Efficient Multimedia summarization system using natural language processing
Pages : 827-831
Download PDF
Abstract
Microblogging offerings have revolutionized the way human beings trade facts. Confronted with the ever-growing numbers of microblogs with multimedia contents and trending topics, it’s far proper to offer visualized summarization to assist users to quickly hold close the essence of topics. While existing works normally attention on text-based strategies best, summarization of a couple of media sorts (e.g., text and image) are scarcely explored. In proposed approach a multimedia microblog summarization framework to automatically generate visualized summaries for trending topics. Specifically, a novel generative probabilistic model, termed multimodalLDA (MMLDA), is proposed to find subtopics from microblogs by means of exploring the correlations amongst different media kinds based on the records accomplished from MMLDA, a multimedia summarizer is designed to one by one pick out representative textual and visual samples and then form a complete visualized summary.
Keywords: Microblog, Summarization, Trending Topic, Social Media, MMLDA