Designing Adaptive Hypermedia for Internet Portals: A Personalization Strategy Featuring Case Base Reasoning with Compositional Adaptation

We propose that the Case Based Reasoning (CBR) paradigm offers an interesting alternative to developing adaptive hypermedia systems, such that the inherent analogy-based reasoning strategy can inductively yield a ‘representative’ user model and the case adaptation techniques can be used for dynamic adaptive personalization of generic hypermedia-based information content. User modeling is achieved by applying CBR retrieval techniques to collect a set of similar past cases, which are used to form a global user-model. Adaptive personalization is accomplished by a compositional adaptation technique that dynamically authors a personalized hypermedia document—a composite of multiple fine-grained information ‘snippets’—by selectively collecting the most relevant information items from matched past cases and systematically amalgamating them to realize a component-based personalized hypermedia document.