<![CDATA[The research aims to develop an accurate and efficient hate speech detection model on Twitter's social media platform by leveraging the power of the Convolutional Neural Network. (CNN). The focus of this research is on identifying hate speeches that are loaded with negative sentiment, especially those related to racial, religious, and sexual orientation issues in the context of the Indonesian language. The research process involved collecting relevant Twitter datasets, preprocessing text to clear and compile data, and word representation using Word2Vec to capture contextual meanings. Specifically designed CNN models are then trained on that dataset. CNN's advantages in automatically extracting semantic features from text, coupled with the use of Word2Vec, allow the model to have high accuracy, which is 87%-99% for emotional assessment and 99% for hate speech assessment. This makes the model very effective in detecting subtle patterns in language that indicate the presence of hate speech. This research has made a significant contribution to the development of a better content moderation system on social media. With its ability to detect hate speech in real time, the model can help create a safer and more inclusive online environment. However, this research still has some limitations, such as limited data set size and variations of hate speech that are not fully represented. Therefore, further research is needed to overcome these limitations and improve the performance of the model.]]>
