Abstract

Social media usage has grown tremendously in the contemporary communication landscape. Along with its numerous benefits, some users abuse the channels by spreading hatred, far from the intended purpose of building connections on a personal level. To date, an empirical method for detecting, quantifying, and categorizing hateful comments on social networks comprehensively and proactively is still lacking. Besides, majority of the cases remain unreported due to social confounders such as fear of victimization and the psychological implications of hateful comments, leading to a situation whereby, the detrimental effect of the situation is underestimated. The ill-defined situation in the growing online space impedes progress towards developing mechanisms and policies to mitigate the harmful effects of hate on social media, ultimately reducing the effectiveness of the platforms as effective communication tools. This proposal suggests Naïve Bayes classifier as a novel approach for detecting and classifying hateful social media comments to bridge this gap. Data set was taken from set provided by Kaggle and consisted of 30,000 Tweets. From the results of the use of this method, it was calculated that Bayes method is 62.75% accurate, which is not satisfactory. However, to bridge accuracy gap, nural algorithm was used which gain an improved accuracy of 87%.

Publication Date

5-2022

Document Type

Master's Project

Student Type

Graduate

Degree Name

Professional Studies (MS)

Department, Program, or Center

Graduate Programs & Research (Dubai)

Advisor

Sanjay Modak

Advisor/Committee Member

Ehsan Warriach

Campus

RIT Dubai

Download

Share

COinS