Arabic Sentiment Analysis based on Neural Network Models: Overview and Comparison

Youssra Zahidi, Yacine El Younoussi, Yassine Al-Amrani

2021

Abstract

Sentiment Analysis (SA) or opinion mining tries to select the sentiment orientation (positive, neutral, or negative) of a text. Arabic Sentiment analysis (ASA) is complicated, it is considered a challenging task compared to other foreign languages by dint of the complications of Arabic at the level of morphology, orthography, its ambiguity, the lack of resources, and various dialects. Deep learning (DL) is a kind of machine learning (ML) and contains several Neural Network (NN) models. The purpose of our work is to debate the issue of DL models that is very important in the ASA domain also provide a comparative analysis of the most valuable and famous NN models that gain salient results in this field, namely: ANN, CNN, RNN, and LSTM. We found through this deep evaluation that the NN models: CNN and LSTM that is a type of RNN have numerous benefits in the ASA field.

Paper Citation

in Bibtex Style

@conference{bml21,
author={Youssra Zahidi and Yacine El Younoussi and Yassine Al-Amrani},
title={Arabic Sentiment Analysis based on Neural Network Models: Overview and Comparison},
booktitle={Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning - Volume 1: BML,},
year={2021},
pages={77-80},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010728700003101},
isbn={978-989-758-559-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning - Volume 1: BML,
TI - Arabic Sentiment Analysis based on Neural Network Models: Overview and Comparison
SN - 978-989-758-559-3
AU - Zahidi Y.
AU - El Younoussi Y.
AU - Al-Amrani Y.
PY - 2021
SP - 77
EP - 80
DO - 10.5220/0010728700003101

in Harvard Style

Zahidi Y., El Younoussi Y. and Al-Amrani Y. (2021). Arabic Sentiment Analysis based on Neural Network Models: Overview and Comparison. In Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning - Volume 1: BML, ISBN 978-989-758-559-3, pages 77-80. DOI: 10.5220/0010728700003101