Classification of pornographic content on Twitter using support vector machine and Naive Bayes

Abstract

The Internet has many benefits, some of them are to gain knowledge and gain the latest information. The internet can be used by anyone and can contain any information, including negative content such as pornographic content, radicalism, racial intolerance, violence, fraud, gambling, security and drugs. Those contents cause the number of children victims of pornography on social media increasing every year. Based on that, it needs a system that detects pornographic content on social media. This study aims to determine the best model to detect the pornographic content. Model selection is determined based on unigram and bigram features, classification algorithm, k-fold cross validation. The classification algorithm used is Support Vector Machine and Naive Bayes. The highest F1-score is yielded by the model with combination of Support Vector Machine, most common words, and combination of unigram and bigram, which returns F1-Score value of 91.14%.