Information Technology Reference
In-Depth Information
Supporting Dynamic, People-Driven Processes
through Self-learning of Message Flows
Christoph Dorn
1
,
2
and Schahram Dustdar
2
1
Institute for Software Research, University of California, Irvine, CA 92697-3455
cdorn@uci.edu
2
Distributed Systems Group, Vienna University of Technology, 1040 Vienna, Austria
lastname@infosys.tuwien.ac.at
Abstract.
Flexibility and automatic learning are key aspects to sup-
port users in dynamic business environments such as value chains across
SMEs or when organizing a large event. Process centric information sys-
tems need to adapt to changing environmental constraints as reflected
in the user's behavior in order to provide suitable activity recommen-
dations. This paper addresses the problem of automatically detecting
and managing message flows in evolving people-driven processes. We in-
troduce a probabilistic process model and message state model to learn
message-activity dependencies, predict message occurrence, and keep the
process model in line with real world user behavior. Our probabilistic pro-
cess engine demonstrates rapid learning of message flow evolution while
maintaining the quality of activity recommendations.
Keywords:
message prediction, process log mining, people-driven
processes, process evolution, message activity dependencies.
1
Introduction
Modern information systems need to enable flexibility and automatic adaptation
capabilities in order to cope with continuously evolving environments where a-
priori fixed requirements are rarely applicable. Organization of multi-national
events such as the Olympic Games or management of value chains across a large
set of Small and Medium-sized Enterprises (SMEs) are just two examples where
exact work practices cannot be precisely defined and executed. In such environ-
ments, users engage in knowledge and coordination intensive workflows that are
subject to continuous change. Processes evolve as participants tune their work
practice to increase eciency and effectiveness. Thus, users want to focus on their
tasks rather than managing and updating their workflow. Instead the involved
information systems should learn from and adapt to the users automatically.
In this paper, we address the case of people-driven dynamic processes. There
exists a process model that describes the activities carried outbyhumansrepre-
senting their expertise. Users are, however, completely free to deviate from the
underlying model which cannot foresee all possible situations. These processes
heavily rely on exchanging unstructured or semi-structured messages such as
