Domain-Driven Approach for Detecting Event Patterns in E-Markets

In this research, we look at the problem of §detecting complex situations arising in e-markets §which have been growing in size rapidly over the §past few years. Existing tools available today still §cannot overcome the main challenges such as §processing a large amount of incoming real-time data §from market feeds and the ability to extract §data efficiently at a high level. We propose to §address these problems using the Event Processing §concept and refer to an e-market as a §distributed event-driven system, e-market §transactions as events and data analysis tools as §Event Processing Systems (EPSs). Our solution is §based on the new EPS architecture that allows the §integration of several existing EPSs (slave EPSs) §under a unified domain-specific user interface and §execution environment. It also includes the unified §e-market data model and event pattern framework for §composing and executing event patterns. Selected §event patterns used for market surveillance in §several stock exchanges have been used in the §evaluation process. We sucessfully developed a §system prototype using two sophisticated commercial §systems, Coral8 and SMARTS, as slave EPSs.