Abstract On social networking sites, online hate speech has become more prevalent due to the quick expansion of mobile computing and Web technology. Previous research has found that being exposed to internet hate speech has substantial offline implications for historically disadvantaged communities. As a result, research into automated hate speech detection has received a lot of interest. Hate speech can have an influence on any population group, but some are more vulnerable than others. An effective automatic hate speech detection ideal, which based on progressive natural language processing and machine learning is not adequate. We need annotated datasets of a size sufficient to train a model. Lack of properly labeled hate speech data, as well as existing biases, have been the biggest obstacles in this field of study for years. To meet these needs, we provide in this paper an unique coral reefs optimization-based method with a transfer learning attitude based BERT (Bidirectional Encoder Representations from Transformers). An optimization approach for coral reefs that simulates coral behaviors for placement and growth in reefs is known as a coral reefs optimization algorithm. In the projected strategy, each solution to the problem is viewed as a coral that is always attempting to be planted and grow in the reefs. Special operators from the coral reefs optimization algorithm are applied to the results at each phase. When all is said and done, the best solution is chosen as the ultimate solution to the issue. A new fine-tuning strategy based on transfer learning is also used to evaluate BERT's ability to capture hateful context in social media posts. To evaluate proposed method, we use datasets that have been annotated for racism, sexism, hate, or objectionable content on Twitter. The results show that our solution outperforms earlier techniques in terms of precision and recall. Keywords: Natural Language Processing; Bidirectional Encoder Representations from Transformers; Coral Reefs Optimization; Hate speech Detection; Twitter.