Hate Speech Detection System using Machine Learning Algorithms

Hateful comments and speeches on social media platforms have emerged as a significant and disturbing issue in recent times. With the rapid modernization of the internet, the proliferation of such content has also increased swiftly. Addressing this problem requires substantial efforts within the sector, particularly in the development of hate speech detection techniques. One effective approach involves the utilization of efficient machine learning models. This paper proposes a model dedicated to the detection of hate speech. The chosen dataset undergoes thorough preprocessing and cleaning, enhancing the quality of the text. Further exploration of the cleaned text aims to provide a more comprehensive understanding. Key features are extracted from the dataset to facilitate model training, incorporating lemmatization, stemming, and the removal of stop words to eliminate unnecessary data. The model is built using CountVectorizer, and various machine learning algorithms are employed to assess performance and gain insights for model improvement. By implementing effective preprocessing techniques and optimizing hyperparameters, our model has demonstrated superior performance, achieving an accuracy of 96%.