The phenomenal growth of social media platforms has resulted in their becoming ubiquitous in the sense that now almost everyone on the planet is using or is being affected by content on social media platforms. Social media platforms have become so influential that they are not only affecting individual thoughts and behaviours but also guiding collective behaviours of groups and societies. There are now innumerable instances of hate speech, abusive content, cyberbullying, misogyny, fake news and disinformation etc. on social media platforms. Such content can severely impact our emotions, mental health, and well-being. The spread of hate speech, misinformation, fundamentalist propaganda, religious hate campaigns etc. on social media platforms can be furthermore dangerous as it could disturb the social order and harmony. The hateful and targeted campaigns can affect social structures and institutions, values, and norms. Therefore, it is extremely important that such content is identified and appropriately dealt with. However, due the huge volume and speed of creation of such content, it can only be done by using sophisticated computational methods that can automatically detect and identify harmful content. Taking into account the fact that the social media is accessible in large number of languages across the world, the task becomes more challenging.
Availability of enough and suitable data and resources is a fundamental requirement towards this endeavour. Asia, being the largest continent, embraces diverse cultures, ethnicities and languages. There are around 2300 languages spoken in Asia. Though there has been substantial research on the abovementioned aspects in the English language, research in Asian languages is still in infancy. The limited or availability of no datasets and resources in these languages is a primary reason for this. This special issue aims to bring together contributions that advance the research in the area of computational methods for automatic detection and identification of harmful content on the social media platforms, such as those reporting:
Algorithmic approaches
Computational resources
Datasets
Dictionaries and Lexicons
Software Resources
Contributions that report novel methods and techniques, datasets and application of various state of the art methods for different tasks in the social media text analytics, including those in low resource languages are also welcome. Though the main focus area of the special issue is on the analysis of the textual content, studies and resources that report multimodal data (with text being the major part) will also be considered.
Topics of Interest
The special issue invites original, unpublished contributions on datasets (elicitation, processing, annotation) and resources (corpora, lexica, database, ontologies, computational approaches, and methodologies) on the following non-exhaustive list of indicative topics:
Aggression and Abusive Content detection
Cognitive Analytics of Social Media Services
Collective Idea Generation and Opinion Dynamics
Depression Intensity Estimation
Detection of Hate Speech, Profanity, Hostility, Cyberbullying
Disinformation, Misinformation, Fake News and Rumours
Emotion analysis, Emotional conversation generation
Fraud detection in online social network
Making online environments safer
Personality trait assessment
Polarization in online discussions
Protecting Children from abusive content
Racial and targeted abuse detection
Religious abuse and bias detection
Sentiment Analysis
Sexism and Misogynistic attitude detection
Social Alignment Contagion in Online Social Networks
Social biases in online texts
Social Perception and Social Influence in social media
Suicide Ideation detection in the Online Environment
Violent Incident detection