Complex Event Processing (CEP) using event-pattern rules has emerged a new paradigm for automated, event-driven decision making. Yet, despite the vast potential CEP offers for agile business applications, its practical relevance rises and falls with the manageability within the organizational framework conditions of an enterprise.<br />In this thesis, we present a novel rule management frame...
Complex Event Processing (CEP) using event-pattern rules has emerged a new paradigm for automated, event-driven decision making. Yet, despite the vast potential CEP offers for agile business applications, its practical relevance rises and falls with the manageability within the organizational framework conditions of an enterprise.<br />In this thesis, we present a novel rule management framework for event-based systems. It caters to the needs of IT experts as well as technically less experienced domain experts through complementary, yet clearly decoupled workflows. IT experts model event-pattern rules in a single, comprehensive model, in parallel and fully integrated with the other elements of an event-processing application. The rule-management workflow for business users, by contrast, builds upon a sophisticated system of configurable building blocks of pattern-detection and reaction logic. Prepared by technically versed domain experts according to the general requirements of an application scenario, these building blocks can be assembled to concrete event-pattern rules by business users in a way that entirely abstracts from underlying complexity.<br /> Other contributions of our work are novel approaches to entity-based state management and hierarchical pattern modeling. Entity-based state management enables the rule-based monitoring of complex, durable entities such as counters, queues, or customer accounts. Hierarchical pattern modeling enables users to compose complex event patterns from smaller, easier-to-understand sub-patterns. Both approaches naturally complement the proposed rule-management framework.<br />The presented concepts are thoroughly evaluated for technical feasibility, applicability, and utility in practical business scenarios.<br />Technical feasibility is demonstrated through their successful implementation as part of the event-processing framework Sense-and-Respond Infrastructure (SARI). Applicability and utility are demonstrated by means of a real-world event-processing application for event-based service assurance and a case study conducted at a leading manufacturer for agricultural machinery.
