Bridging Recommendation and Adaptation: Generic Adaptation Framework - Twittomender compliance case-study

In this paper we consider Recommender System (RS) modeling in terms of Adaptive Hypermedia Systems (AHS) and investigate AHS and RS functionality compliance in terms of common features, functionality, building blocks and composition of the system. We bring up complementary aspects of adaptation, personalization and recommendation in a context of a generic framework which provides properties of information fusion and heterogeneity and could serve as a reference model. We show major recommendation functionality in terms of the reference structure and recommendation process by presenting a conceptual generic ‘adaptation-recommendation’ sequence chart which overlays and combines properties of adaptation and recommendations taking advantages of both. In fact we show that RS if implemented on the web can be considered as AHS, in this wise a generic framework should be capable of describing virtually any RS. In the case study we scrutinize the Twittomender3 RS. We decompose the system in building blocks, outline and highlight its properties along with the advantages and possible enhancements. We conclude by summarizing framework advantages and AH recommendation compliant features as well as lessons learned from this study.