Ulteriori informazioni
Recommender systems (RS) are intended to assist consumers by making choices from a large scope of items. By recommending items with a high likelihood of suiting a consumer¿s needs or preferences, they are able to considerably mitigate the information overload problem at the user¿s side, thus increasing their trust in, satisfaction with, and loyalty to RS providers, such as online shops, internet music catalogs, and online DVD rental services. However, recommendations are prone to errors and often fail to address consumers¿ context specific needs.
Explanations of the underlying reasons behind recommendations can allow users to handle algorithmic errors in recommendations and to better judge their suitability for the users¿ current decision contexts, thus increasing the choice efficiency and effectiveness. The latter, in turn, increases the users¿ acceptance of and satisfaction with RS, as well as it positively affects consumers¿ trust in, loyalty to, and credibility of RS providers. However, in order for these benefits of explanation facilities to surface, they should explain the recommendations in such terms that the consumers themselves use when evaluating their choices.
The latter sets restrictions upon recommendation algorithms constraining them with respect to how recommendations should be produced and what information they should rely on. This interaction between RS and explanation facilities, however, was not covered by recent research on RS.
Therefore, the aim of the current thesis is to narrow this gap and to develop a recommendation technique that accounts for the concurrent objectives of RS, i. e., a method which is capable of providing both accurately predicted recommendations and actionable explanations of the reasons behind them, so that the recommendation process is aligned with the user preference structures.
Sommario
1 Introduction and Motivation
1.1 Motivation
1.2 Objectives
1.3 The Outline of the Thesis
2 Background and Related Work
2.1 A Parsimonious Overview of Recommendation Techniques
2.2 Explanations in Recommender Systems
2.3 Movie-Related Preferences and Relevant Movie Characteristics
2.4 Summary
3 Conceptual Framework of a Hybrid Recommender System that allows for Effective Explanations of Recommendations
3.1 The Modeling of User Preferences
3.2 The Estimation of Model Parameters
3.3 Hybridization with Collaborative Filtering
4 Empirical Study
4.1 The Examined Datasets and Their Properties
4.2 Measures of Prediction Accuracy
4.3 The Employed Algorithms and Benchmarks
4.4 Results
4.5 Summary
5 Conclusions and Future Work
5.1 Research Summary, Key Findings and Contributions
5.2 Discussion and Implications
5.3 Limitations and Avenues for Future Research
Info autore
Paul Marx, born in Novosibirsk (Russia) in 1979, studied Aero-hydrodynamics and Business Administration at the Novosibirsk State Technical University. From 1998 to 2000 he worked as marketing director at Siberian ration supply company "Vital Ltd.". He moved to Germany in 2000, where he continued his studies in Business Administration at the University of Hannover and obtained a master's degree (Dipl.-Ök.) in 2006, specializing in marketing, media research, and economic computer science. At the same time he founded a web surveys service "eQuestionnaire". He subsequently became a lecturer and research assistant at the Chair for Marketing & Media Research at the Bauhaus-University of Weimar, where he completed his doctoral thesis (Dr. rer. pol.) in 2012.
