Abstract:The Business Process Execution Language for Web Services (BPEL) has emerged as the de-facto standard for cross-organizational business processes and business interaction protocols for connecting web services. At the beginning of BPEL, people mainly focused on automated business process execution. Although intended as a language for connecting web services, BPEL has been applied by ambitious people to realize a wide variety of process-aware intra-organizational application, which gives the heavier burden on BPEL than ever. Faced with emerging wider scenarios (complex human activity, various non-web service APIs, etc), original BPEL has difficulties for fulfilling the new requirement. Human activities have complex and dynamic impact on business process execution. BPEL was designed for automated service-to-service interactions, and did not define mechanisms interacting with people in its current version. Considering most modern business processes commonly involving some extent of human-intensive workflow, inability of specification is actually serious, and in some cases fatal. To complement the lack of BPEL on human-driven workflow, and coordinate work performed by human-driven workflows and automated processes, this paper presents a novel four-step procedure. In the first step, we distinguish human activity from automated process among business process, and treat them as external process and internal process, respectively. While the former supports a more event-driven, flexible and dynamic human-driven workflow with unstructured data, the latter, as lower part of the former, supports sequential, predefined and static approach with structured data. In the second step, we utilize State Machine to model the human-driven workflow. The State Machine provides a flexible and efficient way to model the more loosely defined nature of human workflow. In the third step, we transform the State Machine-style workflow to the BPEL-style process. In the fourth step, we integrate external BPEL-style human-driven process with internal sequential automated process into a complete BPEL process. The latter is invoked by the former as a single web service via standard 'invoke' activity in BPEL, or as a sub-process via extensive 'call' activity in BPEL-SPE.

A Static Data Race Detecting Method for BPEL Based on Happens-Before and Lockset

Statically Detect Data Races for WS-BPEL Web Services by Constraint Solver

Finite Capacity Place Method Based Deadlock Prevention Algorithm

Enhanced Data Race Detection Approach Based on Lockset Algorithm

Data Race Detection Tool for Interrupt-Driven Embedded System

Detecting Deadlock in Ada Rendezvous Flow Structure Based on Process Algebra

Detecting Data Races in Interrupt-Driven Programs based on Static Analysis and Dynamic Simulation

Analysis of BPEL4WS Using the Control Dependency Relation Algorithm

Modeling WS-BPEL Process under the Dead-Path-Elimination

Detecting dead statements for concurrent programs

Context-Sensitive Data Race Detection for Concurrent Programs

Approach to detect dead statements for concurrent programs

A New Method to Model and Analyze the WS-BPEL Process under the Dead-Path-Elimination Semantics

A Colored Petri Net Executable Modeling Approach for a Data Flow Well-Structured BPMN Process Model

Detecting Concurrency-Related Problematic Activity Arrangement in WS-BPEL Programs

Optimistic Prediction of Synchronization-Reversal Data Races

Minimal Context-Switching Data Race Detection with Dataflow Tracking

Hybrid Static-Dynamic Analysis of Data Races Caused by Inconsistent Locking Discipline in Device Drivers

A Testing Method for Web Services Composition Based on Data-Flow

Dynamic Recognition of Synchronization Operations for Improved Data Race Detection.

Using State Machine to Integrate Human Activity into BPEL in Dartflow