Hate Speech on Twitter A Pragmatic Approach to Collect Hateful and Offensive Expressions and Perform Hate Speech Detection

Hajime Watanabe, Mondher Bouazizi, Tomoaki Ohtsuki

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

With the rapid growth of social networks and microblogging websites, communication between people from different cultural and psychological backgrounds became more direct, resulting in more and more “cyber” conflicts between these people. Consequently, hate speech is used more and more, to the point where it became a serious problem invading these open spaces. Hate speech refers to the use of aggressive, violent or offensive language, targeting a specific group of people sharing a common property, whether this property is their gender (i.e., sexism), their ethnic group or race (i.e., racism) or their believes and religion, etc. While most of the online social networks and microblogging websites forbid the use of hate speech, the size of these networks and websites makes it almost impossible to control all of their content.Therefore, arises the necessity to detect such speech automatically and filter any content that presents hateful language or language inciting to hatred. In this paper we propose an approach to detect hate expressions on Twitter. Our approach is based on unigrams and patterns that are automatically collected from the training set. These patterns and unigrams are later used, among others, as features to train a machine learning algorithm. Our experiments on a test set composed of 2 010 tweets show that our approach reaches an accuracy equal to 87.4% on detecting whether a tweet is offensive or not (binary classification), and an accuracy equal to 78.4% on detecting whether a tweet is hateful, offensive or clean (ternary classification).

Original languageEnglish
JournalIEEE Access
DOIs
Publication statusAccepted/In press - 2018 Feb 14

Fingerprint

Websites
Learning algorithms
Learning systems
Communication
Experiments

Keywords

  • Hate Speech
  • Machine Learning
  • Sentiment Analysis
  • Twitter

ASJC Scopus subject areas

  • Computer Science(all)
  • Materials Science(all)
  • Engineering(all)

Cite this

Hate Speech on Twitter A Pragmatic Approach to Collect Hateful and Offensive Expressions and Perform Hate Speech Detection. / Watanabe, Hajime; Bouazizi, Mondher; Ohtsuki, Tomoaki.

In: IEEE Access, 14.02.2018.

Research output: Contribution to journalArticle

@article{21a87359ccd94dc0a474b4dc3b5417a9,
title = "Hate Speech on Twitter A Pragmatic Approach to Collect Hateful and Offensive Expressions and Perform Hate Speech Detection",
abstract = "With the rapid growth of social networks and microblogging websites, communication between people from different cultural and psychological backgrounds became more direct, resulting in more and more “cyber” conflicts between these people. Consequently, hate speech is used more and more, to the point where it became a serious problem invading these open spaces. Hate speech refers to the use of aggressive, violent or offensive language, targeting a specific group of people sharing a common property, whether this property is their gender (i.e., sexism), their ethnic group or race (i.e., racism) or their believes and religion, etc. While most of the online social networks and microblogging websites forbid the use of hate speech, the size of these networks and websites makes it almost impossible to control all of their content.Therefore, arises the necessity to detect such speech automatically and filter any content that presents hateful language or language inciting to hatred. In this paper we propose an approach to detect hate expressions on Twitter. Our approach is based on unigrams and patterns that are automatically collected from the training set. These patterns and unigrams are later used, among others, as features to train a machine learning algorithm. Our experiments on a test set composed of 2 010 tweets show that our approach reaches an accuracy equal to 87.4{\%} on detecting whether a tweet is offensive or not (binary classification), and an accuracy equal to 78.4{\%} on detecting whether a tweet is hateful, offensive or clean (ternary classification).",
keywords = "Hate Speech, Machine Learning, Sentiment Analysis, Twitter",
author = "Hajime Watanabe and Mondher Bouazizi and Tomoaki Ohtsuki",
year = "2018",
month = "2",
day = "14",
doi = "10.1109/ACCESS.2018.2806394",
language = "English",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Hate Speech on Twitter A Pragmatic Approach to Collect Hateful and Offensive Expressions and Perform Hate Speech Detection

AU - Watanabe, Hajime

AU - Bouazizi, Mondher

AU - Ohtsuki, Tomoaki

PY - 2018/2/14

Y1 - 2018/2/14

N2 - With the rapid growth of social networks and microblogging websites, communication between people from different cultural and psychological backgrounds became more direct, resulting in more and more “cyber” conflicts between these people. Consequently, hate speech is used more and more, to the point where it became a serious problem invading these open spaces. Hate speech refers to the use of aggressive, violent or offensive language, targeting a specific group of people sharing a common property, whether this property is their gender (i.e., sexism), their ethnic group or race (i.e., racism) or their believes and religion, etc. While most of the online social networks and microblogging websites forbid the use of hate speech, the size of these networks and websites makes it almost impossible to control all of their content.Therefore, arises the necessity to detect such speech automatically and filter any content that presents hateful language or language inciting to hatred. In this paper we propose an approach to detect hate expressions on Twitter. Our approach is based on unigrams and patterns that are automatically collected from the training set. These patterns and unigrams are later used, among others, as features to train a machine learning algorithm. Our experiments on a test set composed of 2 010 tweets show that our approach reaches an accuracy equal to 87.4% on detecting whether a tweet is offensive or not (binary classification), and an accuracy equal to 78.4% on detecting whether a tweet is hateful, offensive or clean (ternary classification).

AB - With the rapid growth of social networks and microblogging websites, communication between people from different cultural and psychological backgrounds became more direct, resulting in more and more “cyber” conflicts between these people. Consequently, hate speech is used more and more, to the point where it became a serious problem invading these open spaces. Hate speech refers to the use of aggressive, violent or offensive language, targeting a specific group of people sharing a common property, whether this property is their gender (i.e., sexism), their ethnic group or race (i.e., racism) or their believes and religion, etc. While most of the online social networks and microblogging websites forbid the use of hate speech, the size of these networks and websites makes it almost impossible to control all of their content.Therefore, arises the necessity to detect such speech automatically and filter any content that presents hateful language or language inciting to hatred. In this paper we propose an approach to detect hate expressions on Twitter. Our approach is based on unigrams and patterns that are automatically collected from the training set. These patterns and unigrams are later used, among others, as features to train a machine learning algorithm. Our experiments on a test set composed of 2 010 tweets show that our approach reaches an accuracy equal to 87.4% on detecting whether a tweet is offensive or not (binary classification), and an accuracy equal to 78.4% on detecting whether a tweet is hateful, offensive or clean (ternary classification).

KW - Hate Speech

KW - Machine Learning

KW - Sentiment Analysis

KW - Twitter

UR - http://www.scopus.com/inward/record.url?scp=85042110053&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85042110053&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2018.2806394

DO - 10.1109/ACCESS.2018.2806394

M3 - Article

AN - SCOPUS:85042110053

JO - IEEE Access

JF - IEEE Access

SN - 2169-3536

ER -