Automated generation of object summaries from relational databases: A novel keyword searching paradigm

Abstract

This paper introduces a novel keyword searching paradigm in relational databases (DBs), where the result of a search is a ranked set of object summaries (OSs). An OS summarizes all data held about a data subject (DS) in the database. More precisely, it is a tree with a tuple containing the keyword as a root and neighboring tuples as children. In contrast to traditional relational keyword search (R-KwS), an OS comprises a more complete and therefore semantically meaningful set of information about the enquired DS. The proposed paradigm is based on two key concepts: Affinity and Importance. The system investigates and quantifies the Affinity of relations in order to automatically create OSs and OS importance (Im(OS)) in order to rank them. Im(OS)s considers the weight (i.e. pagerank) of tuples, Affinity and size of OS. Experimental evaluation on TPC-H and Northwind DBs so far verifies the searching quality of the proposed paradigm.