Journal Article |
Anatomy of Online Hate: Developing a Taxonomy and Machine Learning Models for Identifying and Classifying Hate in Online News Media
View Abstract
Online social media platforms generally attempt to mitigate hateful expressions, as these comments can be detrimental to the health of the community. However, automatically identifying hateful comments can be challenging. We manually label 5,143 hateful expressions posted to YouTube and Facebook videos among a dataset of 137,098 comments from an online news media. We then create a granular taxonomy of different types and targets of online hate and train machine learning models to automatically detect and classify the hateful comments in the full dataset. Our contribution is twofold: 1) creating a granular taxonomy for hateful online comments that includes both types and targets of hateful comments, and 2) experimenting with machine learning, including Logistic Regression, Decision Tree, Random Forest, Adaboost, and Linear SVM, to generate a multiclass, multilabel classification model that automatically detects and categorizes hateful comments in the context of online news media. We find that the best performing model is Linear SVM, with an average F1 score of 0.79 using TF-IDF features. We validate the model by testing its predictive ability, and, relatedly, provide insights on distinct types of hate speech taking place on social media.
|
2018 |
Salminen, J., Almerekhi, H., Milenković, M., Jung, S.G., An, J., Kwak, H. and Jansen, B.J. |
View
Publisher
|
Journal Article |
New Models for Deploying Counterspeech: Measuring Behavioral Change and Sentiment Analysis
View Abstract
The counterterrorism and CVE community has long questioned the effectiveness of counterspeech in countering extremism online. While most evaluation of counterspeech rely on limited reach and engagement metrics, this paper explores two models to better measure behavioral change and sentiment analysis. Conducted via partnerships between Facebook and counter-extremism NGOs, the first model uses A/B testing to analyze the effects of counterspeech exposure on low-prevalence-high-risk audiences engaging with Islamist extremist terrorist content. The second model builds upon online safety intervention approaches and the Redirect Method through a search based “get-help” module, redirecting white-supremacy and Neo-Nazi related search-terms to disengagement NGOs.
|
2021 |
Saltman, E., Kooti, F. and Vockery, K. |
View
Publisher
|
Journal Article |
New Models for Deploying Counterspeech: Measuring Behavioral Change and Sentiment Analysis
View Abstract
The counterterrorism and CVE community has long questioned the effectiveness of counterspeech in countering extremism online. While most evaluation of counterspeech rely on limited reach and engagement metrics, this paper explores two models to better measure behavioral change and sentiment analysis. Conducted via partnerships between Facebook and counter-extremism NGOs, the first model uses A/B testing to analyze the effects of counterspeech exposure on low-prevalence-high-risk audiences engaging with Islamist extremist terrorist content. The second model builds upon online safety intervention approaches and the Redirect Method
through a search based “get-help” module, redirecting whitesupremacy and Neo-Nazi related search-terms to disengagement NGOs.
|
2021 |
Saltman, E., Kooti, F. and Vockery, K. |
View
Publisher
|
Journal Article |
Exploring The Capabilities Of Prevent In Addressing Radicalisation In Cyberspace Within Higher Education
View Abstract
The Counter Terrorism and Security Act (2015) introduced a binding duty on public sector bodies in the United Kingdom (UK), including education, to have ‘due regard to the need to prevent people from being drawn into terrorism’. The Prevent duty has become widely controversial in the Higher Education (HE) sector with questions as to whether it contravenes academic freedom and freedom of speech.
This research seeks to identify how Prevent may be applied to cyberspace to reduce risk of students being radicalised online at universities. Through semi-structured interviews (N= 16) with individuals working in Prevent and HE, attention is given to the capability of monitoring and filtering website content, which must be considered by universities as part of Prevent compliance. In addition, non-technical methods of reducing radicalisation in cyberspace are explored. Consideration is given to building students’ resilience to challenging information they see online through developing counter-narrative content for social media platforms. With students developing counter-narrative content themselves, specifically addressing vulnerability drivers to radicalisation, universities can enhance compliance with Prevent and create counter extremist content which can be used in cyberspace both in and outside of HE.
|
2019 |
Sandford, L. |
View
Publisher
|
Journal Article |
An Italian Twitter Corpus of Hate Speech against Immigrants
View Abstract
The paper describes a recently-created Twitter corpus of about 6,000 tweets, annotated for hate speech against immigrants, and developed to be a reference dataset for an automatic system of hate speech monitoring. The annotation scheme was therefore specifically designed to account for the multiplicity of factors that can contribute to the definition of a hate speech notion, and to offer a broader tagset capable of better representing all those factors, which may increase, or rather mitigate, the impact of the message. This resulted in a scheme that includes, besides hate speech, the following categories: aggressiveness, offensiveness, irony, stereotype, and (on an experimental basis) intensity. The paper hereby presented namely focuses on how this annotation scheme was designed and applied to the corpus. In particular, also comparing the annotation produced by CrowdFlower contributors and by expert annotators, we make some remarks about the value of the novel resource as gold standard, which stems from a preliminary qualitative analysis of the annotated data and on future corpus development.
|
2018 |
Sanguinetti, M., Poletto, F., Bosco, C., Patti, V. and Stranisci, M. |
View
Publisher
|
Journal |
The Internet and Its Potentials for Networking and Identity Seeking: A Study on ISIS
View Abstract
With the accelerating process of globalization and the development of its technological dimension, more and more opportunities and channels are available to the terrorist groups in the world to mobilize resources and advocates. “Islamic State of Iraq and Sham” (ISIS), as the most modern terrorist-excommunicative group (Takfiry), has been able to utilize the Internet and social networks highly adeptly. While ignoring the function of long-term structural and essential factors underlying the formation of ISIS, and also combining the networked society theory and triple forms of identities proposed by Manuel Castells with theoretical discussions on identity making, networking, and mobilization of media, the current article seeks to analyze the role of cyberspace and social networks as accelerating and opportunistic agents in mobilizing resources and disseminating ISIS. Using an explanatory analytical research method, the current article mainly intends to find a reply to the question: What has been the role of online social networks in connection with ISIS as an excommunicative and terrorist group? According to the research hypothesis, due to ISIS’s subtle, prevocational-emotional and targeted utilization of online social networks, the networks have played the role of an accelerator and opportunity maker in some areas including network building, guidance of public opinion, identity making, and the promotion of project identity of this terrorist group. The general conclusion obtained from the article is that ISIS, as the most terrifying and the most modern group equipped with cyber media, has been able to attract many forces out of fanatical religious groups, unemployed people, criminals, etc., worldwide. Additionally, with the recruitment of fanatics, ISIS has been able to accomplish identity making and network building. As a result, regional security and even security in Western countries is also highly endangered.
|
2017 |
Sardarnia, K. |
View
Publisher
|