Detailed Study of Deep Learning Models for Natural Language Processing
Authors:
- 18 December 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 249-253
Abstract
Natural Language Processing involves computational processing, and understanding of human languages. With the increase in computation power, deep learning models are being used for various NLP tasks. Further availability of large datasets of various languages enables the training of deep learning models. Multiple processing layers are used by the deep learning methods for learning representations of data which are hierarchical in nature and which gives excellent results for different NLP tasks. This paper reviews the important models and methods in deep learning which are applied to natural language problems. In particular, Convolutional Neural Network, Recurrent Neural Network, Long Short-Term Memory, Gated Recurrent Unit, Recursive Neural Network have been described. Also, their advantages and suitability to various natural language processing applications such as text classification, sentiment analysis, etc. have been reviewed.Keywords
This publication has 5 references indexed in Scilit:
- An improved Image Description Method Using Recurrent Neural Network with Gated Recurrent Unit2019
- Recent Trends in Deep Learning Based Natural Language Processing [Review Article]IEEE Computational Intelligence Magazine, 2018
- Neural Network Methods for Natural Language ProcessingSynthesis Lectures on Human Language Technologies, 2017
- Deep learning in neural networks: An overviewNeural Networks, 2015
- Natural Language Processing: A Historical Review1994