Representing and Querying Histories of Semistructured Databases Using Multidimensional OEM
Manolis Gergatsoulis, Yannis Stavrakas, Christos Doulkeridis, Vassilis Zafeiris
Abstract:
Multidimensional semistructured data (MSSD) are semistructured
data that present different facets under different contexts.
Context represents alternative worlds, and is expressed by
assigning values to a set of user-defined variables called
dimensions. The notion of context has been incorporated in the
Object Exchange Model (OEM), and the extended model is called
Multidimensional OEM (MOEM), a graph model for MSSD. In this
paper, we explain in detail how MOEM can represent the history of
an OEM database. We discuss how MOEM properties are applied in the
case of representing OEM histories, and show that temporal OEM
snapshots can be obtained from MOEM. We present a system that
implements the proposed ideas, and we use an example scenario to
demonstrate how an underlying MOEM database accommodates changes
in an OEM database. Furthermore, we show that MOEM is capable to
model changes occurring not only in OEM databases, but in
Multidimensional OEM databases as well. The use of
Multidimensional Query Language (MQL), a query language for MSSD,
is proposed for querying the history of OEM databases and MOEM
databases.
Keywords:
Context-Dependent Data, Multidimensional Semistructured Data,
Multidimensional OEM\sep Histories of Semistructured
Databases, Querying Semistructured Database Histories.