Open Access Open Access  Restricted Access Subscription Access

Using Neural Networks Identification of Word Sense

Ms. S. Savitha, Mr. S. Yuvalath

Abstract


Neural Network based methodology recognizes the feeling of the word showing up in a sentence. The distinction implications of a word are recorded as its different faculties in a lexicon. The enhancement in the precision of distinguishing the right word sense will result in better Machine Translation and Information Retrieval Systems. The objective word is tokenized and the distinguished association with different words is resolved utilizing area explicit WordNet. In the proposed framework, we portray a model, which naturally manufactures Neural Networks from definition messages in machine-coherent lexicons, and shows the utilization of these systems for word sense disambiguation. Neural systems are utilized for the development of preparing sets in an administered philosophy. The plan is assessed through exact execution records for two particular spaces to outline its pertinence in the specific situation.


Full Text:

PDF

Refbacks

  • There are currently no refbacks.