Storing semistructured data in relational databases

Abstract

This paper presents an approach to storing semistructured data in relational databases. We focus on semistructured data as extracted from Web pages by a tool called DEBYE (Data Extraction By Example), and organized according to its data model, the DEByE Object Model (DEByEOM). The approach presented here consists in representing the structure of the objects extracted by DEByE by a relational schema and populating the corresponding database accordingly. We also show how to retrieve such objects by automatically transforming high-level query specifications (query patterns) into SQL queries that are executed over the relational database. Experiments results carried out to evaluate our approach are also described.