View article

In 2011 the IBM Computer Watson was beating its human opponents in the American TV quiz show Jeopardy!. However, the questions for the quiz have been developed by a team of human authors. Authoring questions is a difficult task, because in a Jeopardy! game the questions should be neither too easy nor too hard and should fit the general scope of knowledge of the audience and players. Linked Open Data (LOD) provides huge amounts of information that is growing daily. Yet, there is no ranking that determines the importance of LOD facts, as e. g. by querying LOD for movies starring a distinct actor provides numerous answers, whereas it cannot be answered, which of the movies was the most important for this actor. To rank search results for semantic search various heuristics have been developed to cope with the problem of missing rank in the semantic web. This paper proposes a Jeopardy! like quiz …