Related Communities:

Mapping of Graph Data Models into a Canonical Model

Mapping of Graph Data Models into a Canonical Model for the Development of Data Intensive Systems.

Author(s): Stupnikov S. A.
Published:Systems of High Availability. -- Moscow: Radiotechnika, 2014. -- Iss. 2. -- P. 13-31. (In Russian)
The development of methods and facilities for operating of Big Data requires new approaches providing for coping with the diversity of various chaotically developed data models and their languages. For integration of the heterogeneous information resources the unification of their data models is required by their mapping into the canonical information model (serving as the generalized language in the environment of various resource models). Such mappings should preserve information and semantics of their data description language and semantics of the operations of their data manipulation language. This research is devoted to the unification of the graph models the important and fastly developing kind of the existing data models. The essential modern graph models are discussed including their features, application domains, peculiarities of their use under big data manipulation. On the basis of the survey of the existing graph DBMSs the paper concludes that the majority of the existing graph databases are based on the property or attributed graphs in which attributes are assigned to the vertices and edges of the graph. The generalization of such models has been selected in this paper as the source data models to be unified. The issues of proof of information and operation semantics preserving by the mapping of such graph models into the object-frame canonical model applying the formal specification language AMN are considered.
Download: [ Adobe PDF ]

Supported by Synthesis Group