scholarly journals Study of Undersampling Method: Instance Hardness Threshold with Various Estimators for Hate Speech Classification

2018 ◽  
Vol 2 (2) ◽  
Author(s):  
Naufal Azmi Verdikha ◽  
Teguh Bharata Adji ◽  
Adhistya Erna Permanasari
Keyword(s):  
Social Media ◽  
Hate Speech ◽  
Imbalanced Data ◽  
Poor Performance ◽  
Training Data ◽  
Weighting Method ◽  
Imbalanced Data Classification ◽  
Data Problem ◽  
Speech Classification ◽  
Instance Hardness

A text classification system is needed to address the problem of hate speech in social media. However, texts of hate speech are very hard to find in social media. This will make the distribution of training data to be unbalanced (imbalanced data). Classification with imbalanced data will make a poor performance. There are several methods to solve the problem of classification with imbalanced data. One of them is undersampling with Instance Hardness Threshold (IHT) method. IHT method balances the dataset by eliminating data that are frequently misclassified. To find those data, IHT requires an estimator, which is a classifier. This research aims to compare estimators of IHT method to solve imbalanced data problem in hate speech classification using TF-IDF weighting method. This research uses the class ratio of dataset after undersampling, time of the undersampling process, and Index of Balanced Accuracy (IBA) evaluation to determine the best IHT method. The results of this research show that IHT method using the Logistic Regression (IHT(LR)) has the fastest undersampling process (1.91 s), perfectly balance dataset with the class ratio is 1:1, and has the best of IBA evaluation in all estimation process. This result makes IHT(LR) be the best method to solve the imbalanced data problem in hate speech classification.

scholarly journals Hate Speech Classification in Indonesian Language Tweets by Using Convolutional Neural Network

2021 ◽  
Vol 14 (3) ◽  
pp. 225-239
Author(s):  
Dewa Ayu Nadia Taradhita ◽  
I Ketut Gede Darma Putra
Keyword(s):  
Neural Network ◽  
Social Media ◽  
Convolutional Neural Network ◽  
Hate Speech ◽  
Rapid Development ◽  
Term Weighting ◽  
Weighting Method ◽  
Testing Stage ◽  
Testing Accuracy ◽  
Speech Classification

The rapid development of social media, added with the freedom of social media users to express their opinions, has influenced the spread of hate speech aimed at certain groups. Online based hate speech can be identified by the used of derogatory words in social media posts. Various studies on hate speech classification have been done, however, very few researches have been conducted on hate speech classification in the Indonesian language. This paper proposes a convolutional neural network method for classifying hate speech in tweets in the Indonesian language. Datasets for both the training and testing stages were collected from Twitter. The collected tweets were categorized into hate speech and non-hate speech. We used TF-IDF as the term weighting method for feature extraction. The most optimal training accuracy and validation accuracy gained were 90.85% and 88.34% at 45 epochs. For the testing stage, experiments were conducted with different amounts of testing data. The highest testing accuracy was 82.5%, achieved by the dataset with 50 tweets in each category.

scholarly journals Hate Speech Classification in Social Media Using Emotional Analysis

2018 ◽  
Author(s):  
Ricardo Martins ◽  
Marco Gomes ◽  
Jose Joao Almeida ◽  
Paulo Novais ◽  
Pedro Henriques
Keyword(s):  
Social Media ◽  
Hate Speech ◽  
Speech Classification

scholarly journals Imbalanced data classification based on hybrid resampling and twin support vector machine

2017 ◽  
Vol 14 (3) ◽  
pp. 579-595 ◽  
Author(s):  
Lu Cao ◽  
Hong Shen
Keyword(s):  
Support Vector Machine ◽  
Real Life ◽  
Imbalanced Data ◽  
Data Classification ◽  
Training Data ◽  
Twin Support Vector Machine ◽  
Support Vector ◽  
Imbalanced Datasets ◽  
Minority Class ◽  
Imbalanced Data Classification

Imbalanced datasets exist widely in real life. The identification of the minority class in imbalanced datasets tends to be the focus of classification. As a variant of enhanced support vector machine (SVM), the twin support vector machine (TWSVM) provides an effective technique for data classification. TWSVM is based on a relative balance in the training sample dataset and distribution to improve the classification accuracy of the whole dataset, however, it is not effective in dealing with imbalanced data classification problems. In this paper, we propose to combine a re-sampling technique, which utilizes oversampling and under-sampling to balance the training data, with TWSVM to deal with imbalanced data classification. Experimental results show that our proposed approach outperforms other state-of-art methods.

Combating the challenges of social media hate speech in a polarized society

2019 ◽  
Vol 53 (4) ◽  
pp. 501-527
Author(s):  
Collins Udanor ◽  
Chinatu C. Anyanwu
Keyword(s):  
Social Media ◽  
Free Speech ◽  
Hate Speech ◽  
Local Context ◽  
Content Type ◽  
The Social ◽  
Significant Difference ◽  
Word Clouds ◽  
Social Media Platforms ◽  
Speech Classification

Purpose Hate speech in recent times has become a troubling development. It has different meanings to different people in different cultures. The anonymity and ubiquity of the social media provides a breeding ground for hate speech and makes combating it seems like a lost battle. However, what may constitute a hate speech in a cultural or religious neutral society may not be perceived as such in a polarized multi-cultural and multi-religious society like Nigeria. Defining hate speech, therefore, may be contextual. Hate speech in Nigeria may be perceived along ethnic, religious and political boundaries. The purpose of this paper is to check for the presence of hate speech in social media platforms like Twitter, and to what degree is hate speech permissible, if available? It also intends to find out what monitoring mechanisms the social media platforms like Facebook and Twitter have put in place to combat hate speech. Lexalytics is a term coined by the authors from the words lexical analytics for the purpose of opinion mining unstructured texts like tweets. Design/methodology/approach This research developed a Python software called polarized opinions sentiment analyzer (POSA), adopting an ego social network analytics technique in which an individual’s behavior is mined and described. POSA uses a customized Python N-Gram dictionary of local context-based terms that may be considered as hate terms. It then applied the Twitter API to stream tweets from popular and trending Nigerian Twitter handles in politics, ethnicity, religion, social activism, racism, etc., and filtered the tweets against the custom dictionary using unsupervised classification of the texts as either positive or negative sentiments. The outcome is visualized using tables, pie charts and word clouds. A similar implementation was also carried out using R-Studio codes and both results are compared and a t-test was applied to determine if there was a significant difference in the results. The research methodology can be classified as both qualitative and quantitative. Qualitative in terms of data classification, and quantitative in terms of being able to identify the results as either negative or positive from the computation of text to vector. Findings The findings from two sets of experiments on POSA and R are as follows: in the first experiment, the POSA software found that the Twitter handles analyzed contained between 33 and 55 percent hate contents, while the R results show hate contents ranging from 38 to 62 percent. Performing a t-test on both positive and negative scores for both POSA and R-studio, results reveal p-values of 0.389 and 0.289, respectively, on an α value of 0.05, implying that there is no significant difference in the results from POSA and R. During the second experiment performed on 11 local handles with 1,207 tweets, the authors deduce as follows: that the percentage of hate contents classified by POSA is 40 percent, while the percentage of hate contents classified by R is 51 percent. That the accuracy of hate speech classification predicted by POSA is 87 percent, while free speech is 86 percent. And the accuracy of hate speech classification predicted by R is 65 percent, while free speech is 74 percent. This study reveals that neither Twitter nor Facebook has an automated monitoring system for hate speech, and no benchmark is set to decide the level of hate contents allowed in a text. The monitoring is rather done by humans whose assessment is usually subjective and sometimes inconsistent. Research limitations/implications This study establishes the fact that hate speech is on the increase on social media. It also shows that hate mongers can actually be pinned down, with the contents of their messages. The POSA system can be used as a plug-in by Twitter to detect and stop hate speech on its platform. The study was limited to public Twitter handles only. N-grams are effective features for word-sense disambiguation, but when using N-grams, the feature vector could take on enormous proportions and in turn increasing sparsity of the feature vectors. Practical implications The findings of this study show that if urgent measures are not taken to combat hate speech there could be dare consequences, especially in highly polarized societies that are always heated up along religious and ethnic sentiments. On daily basis tempers are flaring in the social media over comments made by participants. This study has also demonstrated that it is possible to implement a technology that can track and terminate hate speech in a micro-blog like Twitter. This can also be extended to other social media platforms. Social implications This study will help to promote a more positive society, ensuring the social media is positively utilized to the benefit of mankind. Originality/value The findings can be used by social media companies to monitor user behaviors, and pin hate crimes to specific persons. Governments and law enforcement bodies can also use the POSA application to track down hate peddlers.

scholarly journals Challenges of Hate Speech Detection in Social Media

2021 ◽  
Vol 2 (2) ◽  
Author(s):  
György Kovács ◽  
Pedro Alonso ◽  
Rajkumar Saini
Keyword(s):  
Social Media ◽  
Language Processing ◽  
Hate Speech ◽  
Automatic Detection ◽  
Training Data ◽  
Speech Detection ◽  
Model Combining ◽  
Increased Risk ◽  
Recent Emergence ◽  
Efficient Learning

AbstractThe detection of hate speech in social media is a crucial task. The uncontrolled spread of hate has the potential to gravely damage our society, and severely harm marginalized people or groups. A major arena for spreading hate speech online is social media. This significantly contributes to the difficulty of automatic detection, as social media posts include paralinguistic signals (e.g. emoticons, and hashtags), and their linguistic content contains plenty of poorly written text. Another difficulty is presented by the context-dependent nature of the task, and the lack of consensus on what constitutes as hate speech, which makes the task difficult even for humans. This makes the task of creating large labeled corpora difficult, and resource consuming. The problem posed by ungrammatical text has been largely mitigated by the recent emergence of deep neural network (DNN) architectures that have the capacity to efficiently learn various features. For this reason, we proposed a deep natural language processing (NLP) model—combining convolutional and recurrent layers—for the automatic detection of hate speech in social media data. We have applied our model on the HASOC2019 corpus, and attained a macro F1 score of 0.63 in hate speech detection on the test set of HASOC. The capacity of DNNs for efficient learning, however, also means an increased risk of overfitting. Particularly, with limited training data available (as was the case for HASOC). For this reason, we investigated different methods for expanding resources used. We have explored various opportunities, such as leveraging unlabeled data, similarly labeled corpora, as well as the use of novel models. Our results showed that by doing so, it was possible to significantly increase the classification score attained.

scholarly journals Lifelong Learning of Hate Speech Classification on Social Media

2021 ◽  
Author(s):  
Jing Qian ◽  
Hong Wang ◽  
Mai ElSherief ◽  
Xifeng Yan
Keyword(s):  
Social Media ◽  
Lifelong Learning ◽  
Hate Speech ◽  
Speech Classification

scholarly journals Fake News (Hoax) Identification on Social Media Twitter using Decision Tree C4.5 Method

2020 ◽  
Vol 4 (4) ◽  
pp. 711-716 ◽  
Author(s):  
Brenda Irena ◽  
Erwin Budi Setiawan
Keyword(s):  
Social Media ◽  
Feature Selection ◽  
Decision Tree ◽  
Test Data ◽  
Information Gain ◽  
Training Data ◽  
Fake News ◽  
Weighting Method ◽  
Exchange Information ◽  
Test Scenarios

Social media is a means to communicate and exchange information between people, and Twitter is one of them. But the information disseminated is not entirely true, but there is some news that is not in accordance with the truth or often called hoaxes. There have been many cases of spreading hoaxes that cause concern and often harm a particular individual or group. So in this research, the authors build a system to identify hoax news on social media Twitter using the Decision Tree C4.5 classification method to the 50,610 tweet data. What distinguishes this research from some researches before is the existence of several test scenarios, classification only, classification using weighting feature, and also classification using weighting feature and feature selection. The weighting method used is TF-IDF, and the feature selection uses Information Gain. The features used are also generated using n-grams consisting of unigram, bigram, and also trigrams. The final results show that the classification test that uses weighting feature and feature selection produces the best accuracy of 72.91% with a ratio of 90% training data and 10% test data (90:10) and the number of features used is 5000 in unigram features.  

scholarly journals Ensemble-based Semi-Supervised Learning for Hate Speech Detection

2021 ◽  
Vol 34 (1) ◽  
Author(s):  
Safa Alsafari
Keyword(s):  
Social Media ◽  
Supervised Learning ◽  
Hate Speech ◽  
Classification Performance ◽  
Media Content ◽  
Learning Approach ◽  
Classification Methods ◽  
Speech Detection ◽  
Speech Classification

Large and accurately labeled textual corpora are vital to developing efficient hate speech classifiers. This paper introduces an ensemble-based semi-supervised learning approach to leverage the availability of abundant social media content. Starting with a reliable hate speech dataset, we train and test diverse classifiers that are then used to label a corpus of one million tweets. Next, we investigate several strategies to select the most confident labels from the obtained pseudo labels. We assess these strategies by re-training all the classifiers with the seed dataset augmented with the trusted pseudo-labeled data. Finally, we demonstrate that our approach improves classification performance over supervised hate speech classification methods.

Sign in / Sign up

Export Citation Format

Share Document