Ruchi: Rating Individual Food Items in Restaurant Reviews

Restaurant recommendation systems are capable of recommending restaurants based on various aspects such as location, facilities and price range. There exists some research that implements restaurant recommendation systems, as well as some famous online recommendation systems such as Yelp. However, automatically rating individual food items of a restaurant based on online customer reviews is an area that has not received much attention. This paper presents Ruchi, a system capable of rating individual food items in restaurants. Ruchi makes use of Named Entity Recognition (NER) techniques to identify food names in restaurant reviews. Typed dependency technique is used to identify opinions associated with different food names in a single sentence, thus it was possible to carry out entity-level sentiment analysis to rate individual food items instead of sentence-level sentiment analysis as done by previous research.