FrameNet-Based Modeling of the Domains of Tourism and Sports for the Development of a Personal Travel Assistant Application

This paper presents an enriched frame-based multilingual lexicon covering the domains of Tourism and Sports, which supports a personal travel assistant application – m.knob – developed to help tourists get recommendations of attractions and activities, as well as to communicate with other tourists and service providers, in the context of major international sports events, such as the Summer Olympics. Recommendations are provided through frame-based automated categorization of tourist attractions based on semantic information extracted from tourists’ comments on online platforms, which are then matched to semantic information extracted from the input the user inserts in a conversational user interface.