Query Pattern-based Relational Data to XML Data Translation Algorithm

Abstract

This paper proposes a new query pattern-based relational schema-to-XML schema translation (QP-T) algorithm to resolve implicit referential integrity issue. Various translation methods have been introduced on structural aspects and/or semantic aspects. However, most of conventional methods consider only explicit referential integrities specified by relational schema. It causes several problems such as incorrect transformation, abnormal relational model transition and so on. Researches about syntactic/semantic structure extraction also have been executed in reverse-engineering part. Many systems or algorithms suggested in Reverse Engineering researches are so complicate and not proper to RDB-to-XML translation. Some of them just consider syntactic structure extraction and others reflect every structure and constraints. That is, most of methods in reverse engineering part include unnecessary part. The QP-T algorithm analyzes query pattern and extract implicit referential integrities through equi-join between columns. The QP-T algorithm was based on a concept that columns related to equi-join in relational schema can have referential integrity. The most distinct contribution of QP-T algorithm was to enhance extraction of referential integrity relation information for translation. Therefore, the QP-T algorithm reflects not only explicit referential integrities but also implicit referential integrities during RDB-to-XML translation.