Despite all the advantages made available by information and communication technology, its applicability is still limited due to problems caused by personal attacks or pseudo-attacks, which are called toxic contents. Since cyberbullying via the usage of toxic digital content on an individual may have severe consequences, it is important to implement various techniques to detect cyberbullying from social media using machine learning approaches. During a cyberbullying detection process, word embedding techniques are used to represent the words for text analysis. Feeding strong word representations to classification methods is an important issue. In this thesis, the effect of ELMo is evaluated against three other word embeddings using various machine learning algorithms. In addition, an ensemble technique based on machine learning models using Elmo word embeddings is proposed to evaluate the effect of ELMo word embedding on the proposed model and compare the results with other base machine learning algorithms.