Social media is a platform for sharing opinions, feelings, advertisements, and more. The widespread availability of free internet access has become accessible to everyone. Research shows that over 90% of individuals using the internet have one or more social media accounts and are actively engaged in sharing their views on trending topics. Machine learning (ML) techniques play a crucial role in processing these reviews. The reviews present in social media of mostly unlabeled i.e., no label is associated with them and thus, the use of traditional parameter accuracy does not work. For these types of reviews analysis clustering approach is preferred. In this work, the Long Short Term Memory (LSTM) approach is used to cluster the reviews into a binary class known as Bi-LSTM. The IMDb dataset with unlabeled reviews is considered for analysis in this work.