This paper presents a method to solve word sense ambiguity using neural network. Most of the previous word sense disambiguation approaches were based on neural networks, having limitations due to their huge feature set size. Here, bigram method is adopted in two ways: post-bigram (hw) and pre-bigram (wn). Two bigram features are treated as input for the networks, each defined with one hidden layer with hidden neurons ranging from two to twenty. The input model is extracted from the sentences in Brown corpus. In this, the performance of the networks are compared using mean squared error values. Among all networks, trainable cascade forward back propagation network gives 71.3% of accuracy with pre-bigram.

Keywords: Neural Network, Word sense disambiguation, post bigram, pre bigram.