Arabic Sentiment Analysis Based on 1-D Convolutional Neural Network

Autor:	Adib Abdellah, Ayad Habib, Bensalah Nouhaila, Ibn El Farouk Abdelhamid
Rok vydání:	2021
Předmět:	Computer science business.industry Deep learning Sentiment analysis Variety (linguistics) computer.software_genre Convolutional neural network Preprocessor Word2vec Artificial intelligence business Arabic sentiment analysis computer Word (computer architecture) Natural language processing
Zdroj:	Innovations in Smart Cities Applications Volume 4 ISBN: 9783030668396
Popis:	In recent years, Deep Learning (DL) has garnered tremendous success in a variety of application domains such as sentiment analysis. In this paper, we study the impact of different preprocessing techniques on Sentiment Analysis (SA) in Arabic. Moreover, we describe the details of selecting good Arabic word embeddings using Word2Vec and FastText models. Furthermore, a DL architecture based on Convolutional Neural Networks (CNNs) is proposed. Experiments on a total of 63K book reviews show the suitability of our proposed system on Arabic SA.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::497930f0a1a437cec0ffafebf4b36cfe https://doi.org/10.1007/978-3-030-66840-2_4 Zobrazit plný text záznamu