Title

Temporal behavioural analysis of extremists on social media: A machine learning based approach

Document Type

Conference Proceeding

Source of Publication

2021 6th International Conference on Smart and Sustainable Technologies, SpliTech 2021

Publication Date

9-8-2021

Abstract

Public opinion is of critical importance to businesses and governments. It represents the collective opinion and prevalent views about a certain topic, policy, or issue. Extreme public opinion consists of extreme views held by individuals that advocate and spread radical ideas for the purpose of radicalizing others. while the proliferation of social media gives unprecedented reach and visibility and a platform for freely expressing public opinion, social media fora can also be used for spreading extreme views, manipulating public opinions, and radicalizing others. In this work, we leverage data mining and analytics techniques to study extreme public opinion expressed using social medial. A dataset of 259, 904 tweets posted between 21/02/2016 and 01/05/2021 was collected in relation to extreme nationalism, hate speech, and supremacy. The collected data was analyzed using a variety to techniques, including sentiment analysis, named entity recognition, social circle analysis, and opinion leaders' identification, and results related to an American politician and an American right-wing activist were presented. The results obtained are very promising and open the door to the ability to monitor the evolution of extreme views and public opinion online.

ISBN

9789532901122

Disciplines

Computer Sciences

Keywords

Extreme views, Nationalism, Opinion leaders, Sentiment analysis, Social media analytics

Scopus ID

85118441623

Indexed in Scopus

yes

Open Access

no

Share

COinS