Towards Summarization of Aggregated Multimedia Verticals Web Search Results

The web is experiencing exponential growth of multimedia content. Due to the enormity of multimedia information, web users employ web search engines as a gateway to access multimedia information. The existing state-of-the-art web search provides access to multimedia content by aggregating media-specific content called verticals. However, these verticals are disjoint and provide multimedia search results in separately ranked linear lists. Moreover, the traditional search results presentation in ranked linear lists poses several challenges to user interaction. Users with exploratory needs have to perform several clicks and scrolls through extensive lists to explore their desired information, hindering efficient interaction with the retrieved information. We propose a generic multimedia verticals summarization framework that aggregates and summarizes the disjoint verticals in conceptually coherent multimedia documents to facilitate the exploration of ranked linear search results. Our proposed framework allows users to perform minimum clicks and scrolls to explore aggregated multimedia search results, hence improving the user search result exploration experience.