The business process is a framework for activities to participate and interact so as to produce a product or a service and achieve the well-defined objectives of an enterprise (Hammer & Champy, 1994; Keung & Kawalek, 1997), and the structure of it greatly affects the overall performance of the enterprise. Moreover, in this ever changing global marketplace, the better the performance of the enterprise, the greater the competitiveness the enterprise can achieve. An enterprise not only faces challenges in external global competition but also faces root challenges of internal business process improvement to drive the performance towards excellence (Schorr, 1998). It is then important to evaluate and analyze the internal business process of the enterprise for the necessary partial or complete redesign of the process to improve the business performance (Morrow & Hazell, 1992).
Very often, an enterprise may need to perform a dynamic analysis of their business process so as to simulate and evaluate different sets of processes that could ensure the efficiency and effectiveness of the business process flow as well as improve the overall performance of the enterprise (Alera, Borrajoa, Camachoa, & Sierra-Alonsob, 2002). Business process analysis is based on the business process flow or business process model that the enterprise has adopted. Business process modeling is one of the business tools that help an enterprise to achieve competitive advantages and improve business performance. The business process flow can be systemically represented and mapped by the model, and business process analysis can then be carried out. With the analyzed results, the business process can be revolutionized, redesigned or reengineered, so the enterprise can then achieve the great benefits of enhanced competitiveness (Evans, Towill, & Naim, 1995).
Business process modeling and business process analysis are two inter-related research areas. In the research area of business process modeling, research interests ranges from process enactment, such as ECA Rules (Bae, Bae, Kang, & Kim, 2004) and collaboration with business process choreography (Jung, Hur, Kang, & Kim, 2004) to process monitoring, such as enterprise information portal (Hur, Bae, & Kang, 2003) and run-time environment (Kim, Kang, Kim, Bae, & Ju, 2000). Some leading approaches to business process modeling representation have been discussed in the literature, such as the family of Integrated Computer Aided Manufacturing Definition (IDEF) (Defining IDEF, 1992, Kusiak et al., 1994 and US Air Force, 1981), petri-nets approach (Van der Aalst & Van Hee, 1996), hyper-graph approach (Huang, 1997), entity relationship modeling (Maker et al., 1992), the role activity diagrams approach (Ould, 1995), and state-driven approach (Lee, Kim, Kang, Kim, & Lee, 2007). On the other hand, research into business process analysis ranges from process analysis to process improvement, entailing aspects such as workflow mining (Van der Aalst & Aj, 2004), process measurement (Cardoso, 2005), and process optimization (Ha, Bae, Park, & Kang, 2006). Moreover, there are a number of analysis techniques that have been discussed in the literature, such as reachability graph (Yang & Liu, 1998), the structural analysis approach (Sadiq & Orlowska, 1999), Queueing theory (Kleinrock, 1975), self-organizing map (SOP) or dimensional output space (Díaz, Domínguez, Cuadrado, & Fuertes, 2008).
The research discussed in the literature and the approaches for business process modeling and analysis all have different advantages and disadvantages. However, research into business process modeling and analysis still has room to improve, since most of the work focuses on the qualitative approach that concerns the logical correctness of the defined process instead of the performance of the defined process. Therefore, in this paper, we propose a quantitative approach using an activity model for business modeling and analysis, in which, adjacent matrixes can be applied to provide explicit performance indicators for the enterprise to identify the inefficient and ineffective activity looping, and the business process flow can then be improved. Moreover, the proposed quantitative methodologies use a time series intervention ARIMA model to measure and compare the simulation results of the business process reengineering. This is based on the process activity analysis, so that the intervention effects and the asymptotic changes can be determined.
After this introduction, the rest of this paper is organized as follows: Section 2 contains the description of the basic six types of business activity interactions. The description of the activity model and analysis for business process are proposed in Section 3. Then a time series intervention ARIMA model for measuring process activity changes is discussed in Section 4. The simulation analysis and the results of time series intervention ARIMA analysis of the business process reengineering based on the proposed activity model analysis are presented in Section 5. Finally, conclusions are given in Section 6.
