Bangla hate speech detection on social media using attention-based recurrent neural network

Hate speech has spread more rapidly through the daily use of technology and, most notably, by sharing your opinions or feelings on social media in a negative aspect. Although numerous works have been carried out in detecting hate speeches in English, German, and other languages, very few works have...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Das Amit Kumar, Al Asif Abdullah, Paul Anik, Hossain Md. Nur
Formato: article
Lenguaje:EN
Publicado: De Gruyter 2021
Materias:
rnn
gru
Q
Acceso en línea:https://doaj.org/article/58b89077a75f4e4788a7dad893e7f789
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:58b89077a75f4e4788a7dad893e7f789
record_format dspace
spelling oai:doaj.org-article:58b89077a75f4e4788a7dad893e7f7892021-12-05T14:10:51ZBangla hate speech detection on social media using attention-based recurrent neural network2191-026X10.1515/jisys-2020-0060https://doaj.org/article/58b89077a75f4e4788a7dad893e7f7892021-04-01T00:00:00Zhttps://doi.org/10.1515/jisys-2020-0060https://doaj.org/toc/2191-026XHate speech has spread more rapidly through the daily use of technology and, most notably, by sharing your opinions or feelings on social media in a negative aspect. Although numerous works have been carried out in detecting hate speeches in English, German, and other languages, very few works have been carried out in the context of the Bengali language. In contrast, millions of people communicate on social media in Bengali. The few existing works that have been carried out need improvements in both accuracy and interpretability. This article proposed encoder–decoder-based machine learning model, a popular tool in NLP, to classify user’s Bengali comments from Facebook pages. A dataset of 7,425 Bengali comments, consisting of seven distinct categories of hate speeches, was used to train and evaluate our model. For extracting and encoding local features from the comments, 1D convolutional layers were used. Finally, the attention mechanism, LSTM, and GRU-based decoders have been used for predicting hate speech categories. Among the three encoder–decoder algorithms, attention-based decoder obtained the best accuracy (77%).Das Amit KumarAl Asif AbdullahPaul AnikHossain Md. NurDe Gruyterarticlernnattention mechanismlstmgrubangla text classificationbangla hate speech detectionScienceQElectronic computers. Computer scienceQA75.5-76.95ENJournal of Intelligent Systems, Vol 30, Iss 1, Pp 578-591 (2021)
institution DOAJ
collection DOAJ
language EN
topic rnn
attention mechanism
lstm
gru
bangla text classification
bangla hate speech detection
Science
Q
Electronic computers. Computer science
QA75.5-76.95
spellingShingle rnn
attention mechanism
lstm
gru
bangla text classification
bangla hate speech detection
Science
Q
Electronic computers. Computer science
QA75.5-76.95
Das Amit Kumar
Al Asif Abdullah
Paul Anik
Hossain Md. Nur
Bangla hate speech detection on social media using attention-based recurrent neural network
description Hate speech has spread more rapidly through the daily use of technology and, most notably, by sharing your opinions or feelings on social media in a negative aspect. Although numerous works have been carried out in detecting hate speeches in English, German, and other languages, very few works have been carried out in the context of the Bengali language. In contrast, millions of people communicate on social media in Bengali. The few existing works that have been carried out need improvements in both accuracy and interpretability. This article proposed encoder–decoder-based machine learning model, a popular tool in NLP, to classify user’s Bengali comments from Facebook pages. A dataset of 7,425 Bengali comments, consisting of seven distinct categories of hate speeches, was used to train and evaluate our model. For extracting and encoding local features from the comments, 1D convolutional layers were used. Finally, the attention mechanism, LSTM, and GRU-based decoders have been used for predicting hate speech categories. Among the three encoder–decoder algorithms, attention-based decoder obtained the best accuracy (77%).
format article
author Das Amit Kumar
Al Asif Abdullah
Paul Anik
Hossain Md. Nur
author_facet Das Amit Kumar
Al Asif Abdullah
Paul Anik
Hossain Md. Nur
author_sort Das Amit Kumar
title Bangla hate speech detection on social media using attention-based recurrent neural network
title_short Bangla hate speech detection on social media using attention-based recurrent neural network
title_full Bangla hate speech detection on social media using attention-based recurrent neural network
title_fullStr Bangla hate speech detection on social media using attention-based recurrent neural network
title_full_unstemmed Bangla hate speech detection on social media using attention-based recurrent neural network
title_sort bangla hate speech detection on social media using attention-based recurrent neural network
publisher De Gruyter
publishDate 2021
url https://doaj.org/article/58b89077a75f4e4788a7dad893e7f789
work_keys_str_mv AT dasamitkumar banglahatespeechdetectiononsocialmediausingattentionbasedrecurrentneuralnetwork
AT alasifabdullah banglahatespeechdetectiononsocialmediausingattentionbasedrecurrentneuralnetwork
AT paulanik banglahatespeechdetectiononsocialmediausingattentionbasedrecurrentneuralnetwork
AT hossainmdnur banglahatespeechdetectiononsocialmediausingattentionbasedrecurrentneuralnetwork
_version_ 1718371667816742912