Discovering and Analyzing Multi-granular Web Search Results

In this paper, we propose an approach based on the use of soft aggregation operators and multi-granular graphs for discovering and analyzing the results of Web searches, organized into granules of distinct resolution. This practice enables user-driven explorations of the topics retrieved in a search process on the Internet, being based on the graphical representation of both the information granules, and their discovered relationships. We further present the application of both the soft operators and multi-granular graphs within the meta-search system Matrioshka, and discuss their semantics and usefulness.