A Heuristic Multimedia Verticals Aggregated Search Approach and User Behavioral Analysis

The web consists of massive multimedia information generated from heterogeneous sources. These sources may include blogs, social media, podcasts, bulletin boards, wikis, etc. Currently, access to this proliferating information on the web is provided by web search engines. The existing web search engines index and organize the content on the web in the form of specialized media-specific repositories c alled verticals. Traditionally, against the user query, four major verticals (web, image, video, and news) are commonly browsed by the users. However, in-depth access to these verticals is provided in disjoint tabs This causes search space contextual switching and hinders the user information-seeking process flow. I n this research, we aggregated the disjoint verticals based on heuristics that the same ranked search results in the verticals may share similarity-based relationships. To test this hypothesis, we hired human experts to rate the similarity score via the standard instruments. We further compare the obtained results with the baseline system, analyze the users information-seeking behavior on the heuristically aggregated verticals approach, and suggest future recommendations based on the results.