With the onset of the new age, consumers are overwhelmed by the many choices of movies that can be watched. A solution gives rise to the need for a recommendation system (RecSys) that will suggest a title to the user based on user requirements. The new RecSys must be able to address the issues that exist within the plethora of general RecSys. In this work, we have implemented a novel Tag-based RecSys to improve the quality and usefulness of recommendations based on user information and movie tags. This paper proposes a bi-clustering approach for simultaneously clustering movies and users. This work is implemented using the MovieLens datasets. In trying to gauge the efficacy of the new RecSys, the final RMSE of our Tag-based RecSys is 0.9984, and the MAE is 0.8450. The system also produces more personalised results for users by personalising scores. The system evaluation shows considerably positive results.