讲座主题： Independence of Data Management and Execution Management for Business Process Management
Title: Independence of Data Management and Execution Management for Business Process Management
Abstract: Business process management (BPM) is vital in enterprise application systems. Workflow management systems (WfMSs) are a critical component of IT and software systems support for BPM. Current WfMSs aim at managing process execution well but take an ad hoc approach in modeling and managing most of the data needed for execution; they do not clearly separate execution management and data management. There are five categories of data to be managed by a WfMS: 1) process specification, business data, execution status, correlations, and resource information. The lack of separation causes serious difficulties in managing data consistency and process change/evolution, and in particular, it is a serious challenge to provide business processes as services (BPaaS).
We illustrate some of the data management weaknesses in WfMSs with concrete examples. We then argue for the need for the "execution independence" principle of cleanly isolating the two at the logical level and thus their respective concerns. To help and facilitate WfMS design to abide by this principle, we formulate the "data self-sufficiency" principle that can be used to guide the logical data design and further introduce the concept of "universal artifacts" that lead to a general approach to an overall data modeling and management framework. Excitingly, universal artifacts bring many interesting problems for both BPM research and development communities. We will explore and discuss three aspects: How does the universal artifact framework work? How to manage data consistency? How can the framework improve BPM and enable new functionality such as BPaaS? It will be clear that there are many issues demanding answers. This presentation is intended to serve as a starting point of the quest for improved and standardized workflow management technology.
Bio: Jianwen Su is Professor of Computer Science at the University of California, Santa Barbara. He received his BS/MS degrees from Fudan University (China) and his PhD degree from the University of Southern California. He held visiting positions at INRIA and Bell Labs, was/is an adjunct professor at Fudan, Peking, and Donghua Universities in China. His research concerns data modeling and query languages, scientific databases, formal verification, web services, and business process management. His current work focuses on modeling and analysis of business processes concerning compositions and management. His work on data with nested structures, incremental query evaluation, constraint databases, web services, and data-centric workflow is widely known and cited. Dr. Su received two IBM Faculty Awards, was a keynote speaker at several international workshops/conferences including ICSOC 2012 and WS-FM 2013. He served/is serving on program committees of many conferences in databases (PODS, ICDT, VLDB, ICDE, EDBT, etc.) and services computing (ICSOC, BPM, WS-FM, ICWE, ICWS, etc.). He was a general co-chair of ICSOC 2013, the general chair of SIGMOD 2001, the PC chair of PODS 2009, and a program co-chair of a few other database/services computing conferences. He served on the Executive Committees of ACM SIGMOD and PODS, editorial boards for IJFCS and JCST, and is an associate editor of IJCIS.