Reasoning from Context

Abstract

The paper describes a new approach of using the Case-Based Reasoning (CBR) methodology in context-aware systems. We elaborate how this technique can be applied to generate recommendations based on the contexts of users respectively objects especially in a mobile scenario. By combining CBR methodology and context awareness a new and powerful way of modelling and reasoning from contexts emerges. Using CBR cases to enclose the context will then enhance the possibilities to compare contexts, determine certain values of context-similarities, and reflect this information in the process of generating recommendations.