Building a Toxic Comments Classification Model
Vara Swetha, R. Anuhya, E. Sai Sowmya, A. Geethanjali · 2021 5th International Conference on Electronics, Communication and Aerospace Technology (ICECA) · 2021
A large portion of online comments are typically beneficial, but a significant portion are harmful in nature. Harmful or toxic remarks are impolite, oppressive, or absurd online remarks that typically cause other clients to leave a conversation. The risk of web-based tormenting and badgering influences the free progression of thoughts by confining individuals’ contradictory assessments. Locales struggle to advance conversations in a viable manner, prompting many networks to limit or shut down client comments entirely. This paper will deliberately analyze the degree of online badgering and characterize the substance into marks to look at the harmfulness as effectively as it could be expected. Here, this research study utilizes six machine learning calculations and apply them to our information to tackle the issue of text classification and distinguish the best machine learning calculation that becomes dependent on our assessment measurements for performing harmful remarks classification. This study targets on analyzing the harmfulness with high accuracy to restrict its unfavorable impacts, which will be a motivation for associations to make fundamental strides.