The DPIL Framework: Tool Support for Agile and Resource-Aware Business Processes?

Abstract

The Declarative Process Intermediate Language (DPIL) is a declarative process modelling language that allows for specifying multiperspective and multi-modal agile, i.e., exible, processes. The expressiveness of DPIL and its suitability for business process modelling have been evaluated with respect to the well-known Work ow Patterns. The DPIL Framework provides a tool set for supporting agile and resourceaware business processes based on the language DPIL. While the DPIL Modeller component is used to create and verify models, the DPIL Navigator depicts a rule-based execution engine for enacting models. It comprises a web-based worklist which allows process participants to choose and perform tasks. In addition to this, the DPIL Miner component allows for the discovery of DPIL models from event logs.

Background and Signi cance to BPM

Two di erent representational paradigms for business processes can be distinguished: procedural models describe which activities can be executed next and declarative, i.e., rule-based models de ne execution constraints that the process has to satisfy [ 1 ]. In exible processes the exact ow of activities cannot be fully determined at design time. These processes heavily depend on human participants, their decisions and expert knowledge. These information cannot be identi ed and formalized in a whole. As a result, these processes require highly exible IT support. Flexible processes are common in healthcare where, e.g., patient diagnosis and treatment processes require exibility to cope with unanticipated circumstances. In brief, a more exible business process and IT support means a greater number of alternative paths. Within a rule-based model initially all paths are considered viable. The more constraints are added to the model the less paths remain. As result, to make a rule-based model more exible constraints have to be removed or weakened. Moreover, a rule-based model focuses on crosscutting relations instead of the ow of activities. Hence, a rule-based approach is well-suited for modelling exible processes [ 1, 2 ]. As mentioned above, declarative modelling provides means for increasing the number of alternative ? Copyright c 2015 for this paper by its authors. Copying permitted for private and academic purposes. use group Student process BusinessTrip { task Apply for trip task Approve Application task Book flight } ensure sequence(Apply for trip, Approve Application) advice "Approval before booking recommended":

roleSequence(Approve Application, Book flight, Student) advice "Flight should be booked by applicant":

binding(Apply for trip, Book flight) milestone "Done": complete(of ApproveApplication) and

complete(of BookFlight) paths without explicitly modelling them. Given the vast amount of alternatives, a di erentiation between mandatory and only recommended actions in the form of modalities is reasonable [ 3 ]. Independent from a speci c modelling paradigm di erent perspectives on a process exist. The organizational perspective, often also referred to as resource perspective deals with the de nition and the allocation of human and non-human resources to activities. Due to the importance of human decision-making and expert knowledge, especially exible processes need to explicitly integrate the organisational perspective [ 2 ]. In many declarative languages, however, an adequate representation of organisational patterns is often still not possible. The Declarative Process Intermediate Language (DPIL) [ 4 ] is a declarative process modelling language that supports the requirements described above. Unlike other declarative languages it is multi-perspective, i.e., it allows for representing several business process perspectives, namely, control ow, data and especially resources. In order to express complex organisational relations, DPIL builds upon a generic organisational meta model. The expressiveness of DPIL and its suitability for business process modelling have been evaluated [ 4 ] with respect to the well-known Work ow Patterns. DPIL supports 52% of the behavioral, 62% of the organizational and 75% of the informational patterns. Like this, DPIL meets around 50% more of the requirements than BPMN does. Moreover, it is multi-modal, meaning that it allows de ning two di erent types of rules: rules representing mandatory relations (called ensure in DPIL) and rules representing recommended relations (called advice in DPIL). The latter are useful, e.g., to re ect good practices. Here, the modeller can give further explanations that are o ered to the user during execution.

DPIL provides a textual notation based on the use of macros to de ne reusable rules. For instance, the sequence macro (sequence(a; b)) states that the existence of a start event of task b implies the previous occurrence of a complete event of task a; and the binding macro (binding(a; b)) states that an activity b is assigned to the same actor that already performed activity a. The roleSequence (roleSequence(a; b; r)) nally depicts that the start of task b by an actor in role r requires the completion of task a before. Fig. 1 shows a model of a simple process for trip management in DPIL. It states that it is mandatory for all applicants to apply for a business trip before it can be approved. Moreover, it is recommended for students that the approval is carried out before a ight is booked. Since the applicant usually knows appointments best it is recommended that she books the ight herself. The DPIL Framework provides a tool set for modelling, executing and analysing exible and resource-aware business processes based on the language DPIL as described above. 2

DPIL Framework: Overview and Demo Guidelines By means of the example process above, the demo will show the functionality of the three modules the framework consists of: the DPIL Modeller, the DPIL Navigator and the DPIL Miner. The framework supports de ning models using a textual editor based on the Eclipse IDE, the DPIL Modeller. The modeling tool support users through syntactic, semantic and qualitative model analysis based on the Xtext framework1. After specifying DPIL models like in Fig. 1 the model is compiled to an executable le that can be enacted by a rule-based execution engine, the DPIL Navigator.

The DPIL Navigator is based on the execution principle as visualized in Fig. 2. Tasks of a DPIL process model undergo a life cycle composed of events that is managed by the engine. A task, e.g., can be started and completed. The current state of a process is then the series of past events (1). Besides the static elements like human tasks, e.g., Apply for trip and Approve application, a process model may specify rules constraining that series of events, e.g., sequence(Apply for trip, Approve application). When the model is executed, the engine simulates one event ahead for every model element (2) and evaluates the resulting series of 1 http://www.eclipse.org/Xtext/ (a) Sequence rule (b) RoleSequence rule (c) Binding rule events on the basis of the rules (3). Each simulated event that does not violate any ensure rule is related to an action that the engine interprets immediately (4). A simulated start of a task by a certain participant, e.g., is interpreted as the assignment of this task to the participant. If the start event violates an advice rule, the action is marked as not recommended. The principle is demonstrated w.r.t. two di erent instances \Business trip to Innsbruck" of the example process. Fig. 3a shows the web-based worklist for a non-student, e.g., a professor. The task Approve Application is not visible since its execution violates the mandatory (ensure) sequence rule. The execution of the other tasks, however, is possible. Fig. 3b depicts the worklist for a student. In case of a user in role Student, the execution of the task Book ight violates the recommended (advice) roleSequence rule. Note, that the execution of the task is still possible, however, the violation of the rule is highlighted through orange colour and the indication text from the model. Fig. 3c shows the task worklist for a third person in the situation where the trip application has already been started by the professor. In this case the execution of Book ight violates the recommended binding rule, i.e., the ight should be booked by the applicant herself.

As well as manually creating a DPIL model using the DPIL modeler, users can also discover models from event logs using the DPIL Miner [ 5 ]. Here, the analyst selects the DPIL macros that should be discovered w.r.t. the provided event log. Fig. 4 shows the user-interface of the DPIL Miner with some discovered organizational patterns. Both the execution engine and the mining module are based on the JBoss Drools platform [ 6 ] that provides a current implementation of the rete rules solver [ 7 ]. Therefore, in both components DPIL rules are transformed to the Drools Rule Language (DRL). 3

Maturity and Future Work

The DPIL Framework is a well-evaluated prototype that is already used for demonstration purposes in academia as well as in industry. It has been developed and used in the KpPQ project. In [ 5 ] we show how it is possible to discover DPIL models from real-life event logs. The DPIL Navigator with the example process deployed is accessible at http://dpilnavigator.kppq.de. Furthermore, di erent screencasts illustrating the di erent features of the DPIL Framework are accessible online: the Modeller at http://modeller.kppq.de, the navigator at http://navigator.kppq.de and the miner at http://miner.kppq.de. Future work will focus on the implementation of DPIL model veri cation techniques and model simulation features that should support users when choosing from a diversity of possible action through look ahead strategies and recommendations.

References

1. D. Fahland , D. Lubke, J. Mendling, H. Reijers , B. Weber , M. Weidlich , and S. Zugal , \ Declarative versus imperative process modeling languages: The issue of understandability," in Enterprise, Business-Process and Information Systems Modeling , pp. 353 { 366 , 2009 .

2. R. Vacul n , R. Hull, T. Heath , C. Cochran , A. Nigam , and P. Sukaviriya , \ Declarative business artifact centric modeling of decision and knowledge intensive business processes," in Enterprise Distributed Object Computing Conference (EDOC), no . Edoc , pp. 151 { 160 , 2011 .

3. G. Regev and A. Wegmann , \ A regulation-based view on business process and supporting system exibility," in Advanced Information Systems Engineering , vol. 5 , pp. 91 { 98 , 2005 .

4. M. Zeising , S. Schonig, and S. Jablonski , \ Towards a Common Platform for the Support of Routine and Agile Business Processes," in Collaborative Computing: Networking, Applications and Worksharing, 2014 .

5. S. Schonig, C. Cabanillas, S. Jablonski , and J. Mendling , \ Mining the Organisational Perspective in Agile Business Processes," in Enterprise, Business-Process and Information Systems Modeling , Springer, 2015 .

6. T. JBoss Drools, \JBoss Drools Documentation - Chapter 7: Rule Language Reference ," 2013 .

7. C. Forgy , \ Rete: A fast algorithm for the many pattern/many object pattern match problem," Arti cial intelligence , vol. 19 , no. 1 , pp. 17 { 37 , 1982 .