| Personalized database systems give users answers tailored to their personal preferences. While numerous preference evaluation methods for databases have been proposed (e.g., skyline, top-k, k dominance, k-frequency), the implementation of these methods at the core of a database system is a double-edged sword. Core implementation provides efficient query processing for arbitrary database queries, however this approach is not practical as each existing (and future) preference method requires a custom query processor implementation. To solve this problem, this thesis proposes FlexPref, a framework for extensible preference evaluation in database systems. FlexPref, implemented in the query processor, aims to support a wide-array of preference evaluation methods in a single extensible code base. Integration with FlexPref is simple, involving the registration of only three functions that capture the essence of the preference method. Once integrated, the preference method "lives" at the core of the database, enabling the efficient execution of preference queries involving common database operations. FlexPref is the cornerstone and core query processing engine of the CareDB context and preference-aware database system, a complete system implemented within PostgreSQL. We describe the architecture and functionality of CareDB. We then describe the technical details of the FlexPref framework. Since FlexPref is implemented in a relational query processing engine, we investigate the theoretical properties of FlexPref within the relational query processing model, focusing on query optimization properties and a theoretical framework that defines the properties a preference method needs to fulfill in order to be supported by FlexPref. To demonstrate the extensibility of FlexPref, we provide case studies showing the implementation of seven state-of-the-art preference evaluation methods within FlexPref. We also introduce a preference query processing framework that gracefully handles dynamic contextual data registered with CareDB (e.g., current traffic, weather) that is expensive to derive at query runtime. We experimentally evaluate all proposed techniques in this thesis using a real system prototype of CareDB implemented in the PostgreSQL open-source database system. |
