OLAP Cubes for Social Searches: Standing on the Shoulders of Giants?

The advent of social bookmarking has signaled a new era towards a more “pluralistic” web, where entities of multiple types (e.g., users, resources, tags) can coexist and get interconnected. In this paper, we propose going beyond classical searches for resources based on keywords to exploring social data starting from any type of entity, i.e., user, resource or annotation, and requesting aggregated views of related entities based on the relationships defined between entities. We map this type of social searching to OLAP query processing bridging the gap between OLAP and the social web and we study various ways to support on-the-fly aggregations of social data. We further describe how data cubes can be used for precomputing and materializing the results of all possible aggregate queries over a social dataset. Our experiments show that cubes can efficiently support searches over social data and they particularly fit searches that combine popular entities over large datasets.