With the modern-day business becoming complex each day, organizations are quickly adapting to various cutting-edge developments that influence the behavior and structure of different business processes. Such processes are the representations of the work including the Business Process Management Systems (BPMS), which offer significant support to the business. As a result, BPMS has undergone a lot of changes or adjustments in its architecture to accommodate the complex processes (Grefen, 2010). Two separate dimensions, including aggregation and abstraction, are used in understanding and exploring the Service Oriented Architecture (SOA) structure. The aggregation dimension determines the granularity while the abstraction dimension is useful in determining the concreteness (Pourmirza et al., 2017). This essay explores how the positioning of architecture along the two dimensions above determine granularity and concreteness.
Granularity refers to the scope or size of the components of the architecture. There are two sub-dimensions of the aggregation dimension, namely (i) the data aggregation and (ii) the system aggregation (Pourmirza et al., 2017). The System-Aggregation Dimension (S-AG) focuses on positioning an architecture in the Level of Elaboration (LoE) cube based on granularity decomposition of the functional components (Pourmirza et al., 2017). On the other hand, Data-Aggregation Dimension (D-AG) positions an architecture in the Level of Detail (LoD) cube. The term, aggregation in architecture, defines the ‘level of detail’ in a given information system concerning the number or scope of the available components (Grefen, 2010). Therefore, high-level aggregation architecture presents a given system as a ‘single black box’. On the contrary, lowest level aggregation architecture indicates systems that have small recognized sub-components.
Figure 1 below is an illustration of the distinction between S-AG (0) – an architecture with a system-aggregation dimension that depicts BPMS as a single black box, and S-AG(1) – an architecture with system-aggregation dimension showing a high-level architectural view of identified components:
Architecture at the First Level and Zero Level of System Aggregation
(Pourmirza et al., 2017)
A similar comparison based on Data-Aggregation Dimension is illustrated in figure 2 below:
Architecture at the First Level and Zero Level of Data Aggregation.
D-AG (0) D-AG (1)
At D-AG (0), architectures do not depict specific data-related component since the entire system is captured as a black box and hence no specification. On the other hand, architectures at D-AG (1) present high-level view of the recognized data-related components of BPMS.
The positioning of abstraction dimension determines the concreteness in a given architecture. According to Pourmirza et al. (2017), highest level abstraction architectures give rough information regarding the components identified while lowest level abstraction architectures present concrete information about the recognized elements. Notably, unlike the highest level abstractions, lowest level abstractions give the final decisions about given products and their versions. There are three distinct abstraction dimension levels namely: (i) AB (1) – architectures of this level show details of very-high-level regarding the components of a given BPMS architecture; (ii) AB (2) – here, the architectures display details of every component of BPMS architecture, which give more concrete information; and (iii) AB (3) – architectures at this abstraction dimension level provide low-level details about the components constituting BPMS architecture. Notably, AB (3) architectures provide some information regarding individual software modules and technologies. Figure 3 below depicts the distinct AB (1) and AB (2) levels.
Components at Abstraction Level 1, 2 and 3
AB (1) AB (2) AB (3)
The first two boxes in figure 2 show the distinction between abstraction level 1 and level 2 while the second two boxes illustrate the distinction between Abstraction level 2 and level 3.
In brief, the positioning of architectures along the aggregation and abstraction dimensions determines the granularity and concreteness respectively. As illustrated in the essay, highest levels of aggregation and abstraction only depict the components of a BPMS system as a black box while lowest levels of aggregation and abstraction present more specific information about the components of a given system.
References
Grefen, P. (2010). Mastering E-Business. London: Routledge.
Pourmirza, S., Peters, S., Dijkman, R., and Grefen, P. (2017). A Systematic Literature Review on the Architecture of Business Process Management Systems. Eindhoven: Eindhoven University of Technology.
