Detection of extreme sentiments on social networks with BERT

Abstract

Online social networking platforms allow people to freely express their ideas, opinions, and emotions negatively or positively. Previous studies have examined sentiments on these platforms to study their behavior in different contexts and purposes. The mechanism of collecting public opinion information has attracted researchers to automatically classify the polarity of public opinions based on the use of concise language in messages, such as tweets, by analyzing social media data. In this paper, we extend the preceding work where an unsupervised approach to automatically detect extreme opinions/posts in social networks is proposed. The performance of the proposed approach is evaluated on five different social network and media datasets. In this work, we use a semi-supervised approach known as BERT to reevaluate the accuracy of our prior approach and the obtained classified dataset. The experiment proves that in these datasets, posts that were previously classified as negative or positive extreme are extremely negative or positive in many cases while using BERT. Furthermore, BERT shows the capability to classify the extreme sentiments when fine-tuned with an appropriate extreme sentiments dataset.

The Relationship Between Government Response Speed and Sentiments of Public Complaints: Empirical Evidence From Big Data on Public Complaints in South Korea

“…At the same time, this article also emphasizes the importance of multidimensional sentiment analysis. Sentiment analysis has been broadly used to classify whether given texts are in negative or positive tones (Cevik et al, 2022; Jamil et al, 2022; J. Kim et al, 2022).…”

Terrorism Attack Classification Using Machine Learning: The Effectiveness of Using Textual Features Extracted from GTD Dataset

“…Pretrained language models have significantly increased in a study recently. Models like BERT have garnered significant attention in NLP research [51]. Accordingly, in this study, the textual features, namely summary attributes, are represented using Distill-BERT.…”

Detection of dangerous events on social media: a critical review

“…A recently published study shows how BERT can efficiently classify extreme negative sentiments in the context of extremism (Jamil et al. 2022 ). CrisisBERT particularly deals with the important task of crisis detection under the classification tasks of crisis detection, and crisis recognition (Liu et al.…”

“…MLM trains a model to predict a random sample of input tokens that have been replaced by a [MASK] placeholder in a multi-class setting over the entire vocabulary (Yamaguchi et al 2021). A recently published study shows how BERT can efficiently classify extreme negative sentiments in the context of extremism (Jamil et al 2022). CrisisBERT particularly deals with the important task of crisis detection under the classification tasks of crisis detection, and crisis recognition (Liu et al 2021).…”