Identifikasi komentar ujaran kebencian pada media sosial indonesia menggunakan metode kombinasi RNN-LSTM untuk identifikasi komentar ujaran kebencian pada data tweet

Hate speech is hate speech or online bullying, repeated acts of attacking, insulting or hurting other people that occur on social media, an example of which is Twitter. Twitter is a medium for exchanging information that is easy to use and very popular. Twitter is also one of the social media that is widely used in Indonesia because it spreads information very quickly. . This research proposes identifying hate speech from Indonesian language social media using a hybrid deep learning approach. Combination of recurrent neural network(RNN) and long short term memory. RNN can map fixed-sized input vector sequences to hidden vector components. LSTM is implemented by considering gradient vector growth components that may exist in the RNN. The steps taken include preprocessing, modeling, implementation, and evaluation. The dataset used comes from. There are 13,158 Indonesian languages, the dataset is from Kaggle, the dataset is divided into two categories, namely tweets and hate speech. The results of the evaluation show that the proposed model produces an f-measure of 86%.