Development of a Collaborative and Constraint-Based Web Configuration System for Personalized Bundling of Products and Services

The composition of product bundles like tourism packages, financial services portfolios or compatible sets of for instance skin care products is a synthesis task that requires knowledgeable information system support. We present a constraint-based Web configurator capable of solving such tasks in e-commerce environments. Our contribution lies in hybridizing a knowledge-based configuration approach with collaborative methods from the domain of recommender systems in order to guide the solving process by user preferences through large product spaces. The system is implemented on the basis of a service-oriented architecture and supports a model-driven approach for knowledge acquisition and maintenance. An evaluation gives acceptable computation times for realistic problem instances.