A uniform representation of multi-variant data in intensive-query databases

In this paper a new approach for the representation of multi-variant data is introduced. Current approaches consist on either hard-core coding techniques or conceptual / logical models to integrate structured and semi-structured data in customized, application-specific ways. The representation introduced here relies instead on unfolding technique to represent multi-variant data uniformly. This leads to a framework with core functionalities for organizing structured and semi-structured data. The paper presents also an efficient methodology towards retrieval of data from the proposed storage along with comparative performance analysis against existing practices. Accuracy, precision, and recall of the proposed technique are quantitatively evaluated and carefully reported.