Showing papers by "Dimitrios Georgakopoulos published in 2006"

From Events to Awareness

Abstract: We define "awareness" as information that is highly relevant to the needs of users and that is delivered in a timely manner. We advocate an event composition approach for computing awareness from various event sources and we describe the Events to Awareness Architecture (E2A) as an embodiment of that approach. E2A is an event-driven publish/subscribe architecture that can be implemented using existing technologies and applied to various domains to provide awareness to users. We describe two different systems where the E2A has been employed to meet the awareness needs of different domains, specifically video surveillance for physical security and human coordination.

Awareness-Enabled Coordination for Large Scale Collaboration Management

Abstract: designed to address the problem of scaling collaboration to large multi-organizational teams. AEC provides a contextualization mechanism that helps its users deal with complex, real world environments where teams involve humans, tools, software services, and agents that come from different organizations, are subject to multiple jurisdictions, and provide diverse expertise. To provide efficiency in achieving team objectives, AEC provides process-based coordination and automation, ongoing policy enforcement, as well as situation and projectrelated awareness. To allow individuals, teams, and organizations to deal with dynamically changing situations, AEC permits dynamic adaptation of user activities, process, resources, policies, organizations, and teams at any time. We use examples from the homeland security and the intelligence gathering domains to illustrate these AEC technical capabilities and their benefits.

Awareness-based Collaboration Driving Process-based Coordination

Abstract: Awareness-enabled coordination (AEC) is a platform designed to address the problem of scaling collaboration to large multi-organizational teams. Such collaboration is inhibited by the complexity in multi-organizational environments and lack of efficiency in achieving team objectives. AEC provides a contextualization mechanism that deals with such complex, real world environments where teams involve humans, tools, software services, and agents that come from different organizations, are subject to multiple jurisdictions, and provide diverse expertise. To provide efficiency in achieving team objectives, AEC provides situation- and project-related awareness, as well as process-based coordination and automation. We describe the AEC architecture and discuss AEC models and mechanisms for computing awareness and coordinating action. We use examples from the homeland security domain to illustrate these AEC technical capabilities and their benefits

Business process learning for real time enterprises

Abstract: Existing approaches for business process mining cannot satisfy Real-Time Enterprise (RTE) goals, such as time-based competition. To support RTE requirements we propose a Process Learning System (PLS) that is capable of learning business processes from a few observed traces and do this in a timeframe that is close to the actual time for completing the process. Unlike existing approaches PLS employs a rich process model that facilitates "guessing" business processes, utilizes domain-specific knowledge captured by activity and resource ontologies, ensures that learned processes comply with specified business rules, and optimizes them to reduce required cost and time. In this paper we focus on the architecture of PLS, and describe the functionality and algorithms employed by key PLS components. We use examples from initial experiments involving learning of processes that assemble complex products from specialized parts.

Software architecture for flexible integration of process model synthesis methods

Abstract: In this paper we suggest an architecture that would integrate various methods for synthesis of a software process model based on domain knowledge about artifacts, process fragments, tools, and limited process execution observations. Our approach suggests using a meta-process specification for integration of various process synthesis methods to provide a generalized process model. We also propose using a process execution observation for confirmation of a synthesized process model.