Big-data Integration Methodologies for effective management and data mining of petroleum digital ecosystems

Abstract

Petroleum industries' big data characterize heterogeneity and they are often multidimensional in nature. In the recent past, explorers narrate petroleum system, as an ecosystem, in which elements and processes are constantly interacted and communicated each other. Exploration is one of the key super-type data dimensions of petroleum ecosystem, (including seismic dimension), exhibiting high degree of heterogeneity, sequence identity and structural similarity; this is especially the case for, elements and processes that are unique to petroleum systems of South East Asia. Existing approaches of petroleum data organizations have limitations in capturing and integrating petroleum systems data. An alternative method uses ontologies and does not rely on keywords or similarity metrics. The conceptual framework of petroleum ontology (PO) is to promote reuse of concepts and a set of algebraic operators for querying petroleum ontology instances. This ontology-based fine-grained multidimensional data structuring adapts to warehouse metadata modeling. The data integration process facilitates to metadata models, which are deduced for Indonesian sedimentary basins, and is useful for data mining and subsequent data interpretation including geological knowledge mapping.