ISRE-Framework: Nonlinear and Multimodal Exploration of Image Search Result Spaces

The extensive information delivery power and an immense volume of image objects make them frequently use multimedia content over the web. However, access to desired image objects to satisfy visual information needs by employing primitive exploration paradigms is difficult. Traditionally, the linear presentation of web image results often leads to reachability and navigation issues. Alternatively, nonlinear approaches provide navigation in web image results. The in-depth browsing to access particular web image results is challenging. In this research, we proposed an exploration framework to browse and explore web image results. We addressed the associated exploration issues, i.e., reachability and navigation in browsing and visualization. The framework enables the nonlinear and multimodal exploration of web image results by representing them in a graph-cluster data model and enabling an interactive exploration mechanism. The graph-cluster data model mainly employs and modifies Zahn’s method and particular algorithms to transform the web image results into specific nonlinear and multimodal search results spaces. The exploration mechanism enables reachability, navigation, browsing, and visualization of web image results in an integrated way. We instantiated the proposed framework over a real dataset of image objects and employed empirical, usability, and comparison tests to evaluate the proposed exploration framework.