Extending the Object-Process Methodology to Big Data Systems

Abstract

There is a general consensus in the software literature that big data systems are difficult to model, specify, and design. It is an important and challenging task to develop an intuitive and easy-to-use, yet coherent and concise method for specifying such systems. The Object-Process Methodology (OPM) graphically specifies systems in an integrated model that describes the staticstructural and behavioural-procedural aspects of a system using Object-Process Diagrams (OPD), we present OPM/D as an extension of OPM for specification of big data systems. Focusing on the four phases of the value of big data, we extend OPM in 4 points: data description, time constraints, concurrency management and status management. A detailed example denotes the accuracy, simplicity and expressiveness of OPM/D to express big data applications.