Solution Architecture for Visitor Segmentation and Recommendation Generation in Real Time

With increasing product portfolios of eCommerce companies it is getting harder for their customers to find the products they like best. A solution to this problem is to analyze the customer's behavior, and recommend products based on ratings. By considering click-stream data, the customer is unburdened with explicitly rating his favored. In this paper, we introduce a system for segmenting visitors and recommending adequate items in real time by using an event-based system called SARI (Sense and Respond Infrastructure) for processing click-stream data. We present the architecture and components for a real-time click-stream analysis which can be easily customized to business needs of domain experts and business users. SARI provides functionality to monitor visitor and customer behavior, respond accordingly and at the same time optimize and adapt customer processes in real time. To illustrate this approach, we introduce a reference implementation, its underlying infrastructure and business scenarios.