Abstract:
We frequently come into negative behaviors on the internet, despite all of its advantages, such as emotional abuse, hate speech, malicious communication, cyberbullying, and cyberthreats. Hate speech tends to become more prevalent online with the current increase in internet usage. Tamil is a low-resource language, hence there aren't many barriers to stop the offences on most platforms. Researching Tamil comments and texts is more challenging because of the diversity of formats and literary styles in which they might be written. This study's primary goal is to identify and track hate speech. The selected approach is made after weighing several conflicting viewpoints to find a single, trustworthy answer. The author advises identifying the unique writing styles of Sinhala texts before attempting to identify hate speech. Unlike most other research on this topic of hate speech, this study's methodology goes beyond just looking for terms that are prohibited. The author is working with deep learning and machine learning approaches in this strategy.
