Information systems architectures are becoming increasingly complex and fragmented. As a result, organizations struggle to cope with change propagation, compliance management, and interoperability. Two major components in current information system architectures in terms of modeling business requirements are business processes and business rules. Although the need for business processes and business rules to be modeled in an integrated manner is well established, the body of knowledge on integrated modeling of the two is limited. To investigate in this topic, the Ph.D. project is divided into 4 studies. The first study aims to identify and evaluate what factors affect integrated vs separated modeling decisions. The second study aims to examine the effect of integrated modeling on process understanding. The third study aims to develop a decision framework that can guide business process modelers in making decisions about how to model a business rule.
The modeling of business processes and business rules has been an important topic of
Information Systems and Computer Science research over the last two decades [
We argue, along the lines of [
In addition to understanding the factors that affect the business rule modeling, it is
also important to understand the effect of separated vs integrated modeling on the
model user. Researchers have argued that model integration leads to improved
process understanding, better communication, and system design, increased
interoperability capacity, and better change propagation of new requirements and increased
capacity for compliance management [
Understanding the factors that business rule modeling decisions and the effect of integrated modeling on the user then provides input in the development of a decision framework. Thus, the aim of the third study is to develop and evaluate a decision framework that guides modelers on whether to integrate a business rule into a business process model, based on research results from the first two studies.
To identify the factors that affect business rule modeling decisions, we conducted a
systematic literature review based on a comprehensive set of well-regarded
Information Systems and Computer Science journals and conferences (see www.aisnet.org
and www.core.edu.au). Our data set consisted of 43,021 full-text articles (see [
To evaluate the relevance of the identified factors and investigate their relative importance, we conducted an empirical evaluation with the authors of the 78 papers relevant for the factor identification being the target participants. The survey was designed, pilot-tested and revised through two iterations. With our finalized survey instrument, we collected 1) the importance of factors, 2) an importance ranking of the factors from each participant and 3) expert opinions on how rules should be modeled given each factor. We sent invitations to 112 authors of the 78 papers and received 22 usable responses, which represents a response rate of 23.08% when calculated as responses per paper. 2.2
In total, twelve factors were identified. For further discussion of the factors, and the
sources/papers in which they were identified, please refer to [
The evaluation consists of two parts. The first part refers to factor importance, and
the second part refers to business rule modeling decisions. To distinguish the relative
importance of each factor, we asked the participants to select at least 5 most important
factors and rank them according to their relative importance. To calculate consensus
between the participants, the rankings provided by all participants are aggregated into
a single score. We adopted the classical positional Borda’s method [
As shown in Table 1, agility is ranked as the most important factor, with 42 points, and criticality is a close second. The factors rate of change and reusability are jointly ranked third, with 37 points. Accessibility, awareness of impact, complexity, governance responsibility and scope of impact follow in that order. The lowest ranked three factors are found to be those of aspect of change, implementation responsibility, and rule source.
While Borda’s method allows us to identify the relative ranking, it is important to determine whether there is an adequate level of agreement between experts’ individual rankings. The concordance of the rankings is an indicator of such agreement. We use compactness [18], to calculate the degree of agreement as suggested in [19]. The compactness of all the rankings is 0.36, resulting the degree of agreement among the participants' rankings is 0.64, which is deemed acceptable [19]. Table 1 also shows the standard deviation for each factor to provide an indication of the level of agreement on that factor.
In terms of the decision analysis, we first distinguish between ‘affecting’ factors and ‘non-affecting’ factors, i.e. no significant difference in expert opinion as to how that factor affects modeling, then analyze the affecting factors to determine modeling guidance given the factors’ circumstances (see [17]). The modeling decision is analyzed for each circumstance of a given factor. Modeling guidance can be derived for the following seven situations: 1. When a rule has relatively high agility, it should be modeled independently. 2. When a rule changes frequently, it should be modeled independently. 3. When a rule changes infrequently it should be integrated into a business process model. 4. When a rule is highly reusable, it should be modeled independently. 5. When a rule's reusability is low, it should be integrated into a business process model. 6. When a rule requires relatively high accessibility, it should be modeled independently. 7. When a rule comes from an external source, it should be modeled independently. 3 3.1
We look to existing theories of cognition science and information representation (for example [20–22]) to understand the effects of integrating business process models and business rules on user understanding of the models. The key arguments are as follows: (1) Due to the limit of working memory capacity and cognitive resources, a heavy cognitive load or cognitive overload typically creates errors, and the rate of error increases with the level of cognitive load [23]. (2) The form of information representation significantly affects cognitive load [21, 24]. (3) Static pictures and diagrams are more comprehensive and easier to make inference than sentential representations in terms of information explicitness and search efficiency [21]. (4) Information presented in an integrated manner is considered to reduce cognitive load, while splitsource information can generate a heavy cognitive load in the process of information assimilation [22].
We introduce a cognitive process in the context of integrated process modeling as consisting of four stages, viz. rule awareness, rule locating, rule comprehension, and information integration. The stages are derived from a human information searching and processing cognitive model, where the information processing are in 5 stages, viz. goal formation, category selection, information extraction, integration, and recycling [25, 26]. In the following, we outline each of the four stages and provide related argumentation based on cognitive load and information representation theories.
Rule awareness: Researchers have found that it is a basic human cognition feature to be aware of information if indications of relevance are explicitly provided [27], and diagram, by its nature, can explicitly connect relevant elements together by placing the elements at adjacent locations, or by associate the elements using a variety types of lines [21]. Hence, we argue that it is easier to be aware of a business rule if explicitly integrated into a business process model.
Rule locating: By integrating a rule into a process model using link information, a stakeholder is able to find the correct business rules without searching the rule list comprehensively thus save cognitive efforts and avoid mistakes. Representing business rules as text annotations or diagrammatically will not require the effort of locating a business rule, since activities and rules are connected with association lines, or organized at adjacent locations.
Rule comprehension: Business process modeling languages generally have simple syntax and semantics, while business rules languages are often abstract, and have a logical syntax which requires some expertise for interpretation and modeling [28]. We argue that business rules integrated into business process models, using graphical constructs, can be better comprehended.
Information integration: By inserting a rule in an appropriate location on the process model and incorporating a rule and an activity into a single element by connections and links, the cognitive load of splitting attention, cross-referencing, and mentally information integration of different information sources are not required. Moreover, the explicit relations between rules and activities in an integrated graphical representation are able to map onto the relations between the features of the process being modeled in such a way that they restrict or enforce the kinds of interpretations that can be made, which makes perceptual inferences extremely easy. 3.2
We hypothesize that the integration of business rules into business process models can improve the understanding of business processes. To empirically validate our hypothesis, we have designed an experiment with participants who have been trained in business process modeling in their courses. The subjects will be divided into two groups. A business process model and a set of relevant business rules will be given to the two groups, in an integrated manner and a separated manner respectively. In terms of measurements, we anticipate that besides traditional understanding performance measurements such as time to complete task and number of errors made, which only provide data on the overall performance, measurements that capture the process of cognition are essential. We will use eye-tracking devices, which can collect a variety of cognitive behavior data, such as eye-fixations, attentional switching, and scan path similarity, to explore empirically the effect of integrated and separated modeling of business processes and rules. This experiment is expected to be conducted in mid2016. 4
The design of the decision framework is still in its early stages. The decision framework is intended to guide process modelers in making informed decisions regarding which rules to integrate and how. The implementation of the decision framework will be through a web application. Before the actual use of the web application, a functionality and usability test of the decision framework will be carried out.
The evaluation of the decision framework includes experimental evaluation and empirical evaluation. The approach of how to evaluate the usefulness of the decision framework is difficult to finalize without finalizing the decision framework. However, it is expected that an experimental method will be appropriate. We will reach out to 35 organizations that have a suitable environment for business requirements modeling. Modelers in these organizations will then be invited to work guided by the framework through a series of activities that will collect both perceptionary as well as real data on use. We will then analyze this data to evaluate the usefulness of the decision framework for improving both the work practice of the business users as well as mitigating the disparity of business requirements modeling for the organization. 5
Following design science, my Ph.D. thesis will develop a decision framework and will contribute to practical knowledge on business process and rule modeling, conceptual modeling theory, information representation theory, and decision support systems. Expected contributions include: 1) practical knowledge of what factors impact integrated/separated business rule modeling decisions, 2) effect of business rule integration on process model understanding, and 3) a decision framework to guide modelers in regards to integrated vs separated rule modeling. 15. Dwork, C., Kumar, R., Naor, M., Sivakumar, D.: Rank aggregation methods for the web.
In: Proceedings of the 10th international conference on World Wide Web. pp. 613–622.
ACM (2001). 16. Young, H.P.: An axiomatization of Borda’s rule. Journal of Economic Theory. 9, 43–52 (1974). 17. Wang, W., Indulska, M., Sadiq, S.: To Integrate or not to Integrate – The Business Rules Question. In: Proceedings of the 28th Internation Conference on Advanced Information Systems Engineering. Springer (2016). 18. Xiaoyun, C., Yi, C., Xiaoli, Q., Min, Y., Yanshan, H.: PGMCLU: A novel parallel gridbased clustering algorithm for multi-density datasets. In: 1st IEEE Symposium on Web Society, 2009. SWS ’09. pp. 166–171 (2009). 19. Chen, L., Li, X., Han, J.: Medrank: discovering influential medical treatments from literature by information network analysis. In: Proceedings of the Twenty-Fourth Australasian Database Conference-Volume 137. pp. 3–12. Australian Computer Society, Inc. (2013). 20. Chandler, P., Sweller, J.: Cognitive load theory and the format of instruction. Cognition and instruction. 8, 293–332 (1991). 21. Larkin, J.H., Simon, H.A.: Why a diagram is (sometimes) worth ten thousand words. Cognitive science. 11, 65–100 (1987). 22. Chandler, P., Sweller, J.: The split-attention effect as a factor in the design of instruction.
British Journal of Educational Psychology. 62, 233–246 (1992). 23. Mousavi, S.Y., Low, R., Sweller, J.: Reducing cognitive load by mixing auditory and visual presentation modes. Journal of educational psychology. 87, 319 (1995). 24. Scaife, M., Rogers, Y.: External cognition: how do graphical representations work? International Journal of Human-Computer Studies. 45, 185–213 (1996). 25. Dreher, M.J., Guthrie, J.T.: Cognitive Processes in Textbook Chapter Search Tasks. Reading Research Quarterly. 25, 323–339 (1990). 26. Guthrie, J.T.: Locating Information in Documents: Examination of a Cognitive Model.
Reading Research Quarterly. 23, 178–199 (1988). 27. Sperber, D., Wilson, D., Ziran He, Yongping Ran: Relevance: Communication and cognition. Citeseer (1986). 28. Lu, R., Sadiq, S.: A survey of comparative business process modeling approaches. In: Business Information Systems. pp. 82–94. Springer (2007).