NER TAGGER

Named Entity Recognition (NER) is the task of identifying and classifying the rigid designators such as person names, place names, organization names etc, in a given document. NER can be visualized as a sequence labeling task. The system give here will work for General Domain and also for Electronic products domain.

Ontology

This work is an attempt to develop an ontology for Natural Language Processing in which all the nouns could be represented in an hierarchy. The ontology we use is a language ontology derived from the subcategorization features. The subcategorization features explain the nature of the noun. It includes the features such as [animate], [concrete], [edible] etc. When these features are assigned to nouns, in a sentence, we get more semantic information about the noun in that sentence. The subcategorization features of the noun ``airplane" give the characteristics of airplane. It is a non-living entity, physically existing, solid, man-made object. This is a vehicle, which has wheels, and it can fly. The features below give these characteristics


Airplane:{-living, +concrete, +movable,}
              {+artifact, +solid, +instrument, +vehicle,}
              {+wheeled, +avion}

The subcategorization features of the abstract noun ``pain" is given below. This is not a physically existing entity, and this is not virtual. This is a feeling which can be sensed.

Pain:{-living, -concrete, -virtual,}
       {-feature, +sensible, +feeling}