Named entity classification using dependency grammar

Named Entity (NE) classification is a task to classify words or group of words in a sentence into some predefined classes like, Person, Organization, Location etc. It is used in different natural language processing (NLP) tasks like machine translation, information retrieval, text annotation, question answering, summarization etc. To classify the Named Entity, researchers use different techniques like rule based entity sorting, statistical class modeling, linear chaining etc. However, most of those techniques do not capture grammatical dependencies of words hidden in the sentence, which is very essential to classify a Named Entity. In our research, we consider grammatical dependencies of words hidden in the sentence and use them to learn a classification model which classifies the Named Entity. In a classification task among three different classes, average performance gain of the proposed system over the state of the art system exceeds more than 6% in F1 Score.