Abstract

Morphological declension, which aims to inflect nouns to indicate number,case and gender, is an important task in natural language processing (NLP).This research proposal seeks to address the degree to which Recurrent NeuralNetworks (RNNs) are efficient in learning to decline noun cases. Given thechallenge of data sparsity in processing morphologically rich languages andalso, the flexibility of sentence structures in such languages, we believe thatmodeling morphological dependencies can improve the performance of neuralnetwork models. It is suggested to carry out various experiments to understandthe interpretable features that may lead to a better generalization of thelearned models on cross-lingual tasks.

 

Quick Read (beta)