Efficient Processing of Preference Queries in Distributed and Spatial Databases
MetadataShow full item record
Traditional SQL queries are recognized for producing an exact and complete result set. However, for an increasing number of applications that manage massive amounts of data, the large result set produced by traditional SQL queries has become difficult to handle. Therefore, there is an increasing interest in queries that produce a more concise result set. Preference queries capture the wishes of the users to produce a result set containing only the most important objects. A naive way of processing a preference query is first evaluating all objects, and then selecting the best ones. This approach is prohibitively costly and does not scale for large datasets. In this thesis, we focus on efficient processing of preference queries in spatial and distributed databases. We propose novel techniques that improve the performance of preference queries avoiding evaluating all objects. The main contributions are in the efficient processing of the following types of preference queries: • Skyline queries in distributed systems. • Spatial preference queries. • Top-k spatial keyword queries. • Top-k spatial keyword queries on road networks. The approaches we propose have been validated through extensive experiments employing real and synthetic datasets. The results we obtained are promising and show the efficiency of our approaches in improving the performance of preference queries in distributed and spatial databases.