Towards intelligent query processing based on Attribute-Oriented Generalization

Due to the intrinsic characteristics of the relational model, the standard database user interface forces users to be familiar with the database schema and underlying data to improve the efficiency of information retrieval. It proposes an intelligent query processing based on Attribute-Oriented Generalization (AOG) to better user experience during the information retrieval process. Firstly, it adds semantics discarded by the relational model to raw data through attribute-oriented generalization, and builds instances of Type Abstraction Hierarchy (TAH). Secondly, it constructs the knowledge base, which is designed on a relational mode, so that both the knowledge base and the underlying relational database can be handled in a single formalism by a relational query language. Thirdly, with the derived specific knowledge base incorporated into the underlying database, a prototype intelligent, interactive and straightforward query processing system with B/S architecture has been built at the top of SQL, which returns the semantically neighboring values and higher level more abstract values through the instance of TAH according to the results of the interaction with the user. Finally, three practical query examples are presented to further exemplify the main ideas and demonstrate the usefulness of the proposed query answering processes.