Event Mining in Virtual Markets Market Research, Knowledge Engineering, Information Agents, and Social Role Structures

We present a generic framework for event mining in e-business. The framework supports customer profiling, the measurement of market dynamics, the evaluation of Web-Sites, the composition of market knowledge, the design of software-agents based on the context/subtext paradigm in the performing arts, and the deduction of social role structures for information societies. And it generalises to a wide range of further application scenarios. The framework is based on the deduction of correlations between a priori observables and a posteriori observables of events and it is intended for the identification of statistical patterns in the event streams of virtual markets. Technically, it can be implemented with tracing facilities for request events in distributed computing environments. We first present the framework and formal representation and modelling techniques. Then we discuss applications. And finally we discuss the possible role of the framework as a boundary object for interdisciplinary research