A Browsing Approach to Explore Web Image Search Results

Image objects are frequently used online available multimedia content due to their extensive information delivering power. The rapid development of image search engines fulfills the user's image information needs. Traditionally the unsatisfactory linear presentations of image results often lead issues mainly related to browsing, exploration, and visualization. The exploration of image results allows us to navigate, organize, and compare the results. Navigation mainly involves browsing, which will enable users to find the desired image results satisfactorily. Due to the linear presentation and browsing approaches, the current image results exploration is inefficient. In this research, an approach is proposed to browse and explore image results while reducing the associated reachability issues. The approach utilizes the multimodal graph for non-linear exploration of web image results. The similarity relationships in the graph are further exploited to partition the groups into multiple clusters, whereas multimodal similarity relationship deals with textual and visual information associated with image results. Moreover, a search tool is also built to provide an interactive search user interface design to explore the image results in a usable way.