meal&me : Leveraging Semantic Technologies for Personalized Restaurant Recommendation

Davide Buscaldi, Hamid Hammouche, Didier Cholvy

Keywords: Recommender systems, ontologies, knowledge engineering

TL;DR: Extending a tourism ontology to enable semantic web technologies for a restaurant recommendation app

Abstract: meal&me is a restaurant recommendation service based on Natural LanguageProcessing (NLP) guided analysis of lived experiences in publicly accessible reviews. The innovation of meal&me consists in the fact that the user experiences are analyzed at a fine-grained level, taking into account real experiences in restaurants and matched to users’ implicit wishes. The integration of a semantic understanding of customer experiences is, in our opinion, the key to making highly personalized recommendations. To implement this semantic understand-ing we applied an innovative method for aspect-based sentiment analysis and knowledge engineering.