An Intelligent Context Based Movie Recommendation System Using Bi-Directional Gated Recurrent Unit

In recent years, there is the rapid expansion of digital content and streaming platforms. There are several existing recommendation systems developed based on collaborative filtering and content-based filtering but still they face challenges such as privacy concerns, overcoming the cold-start problem for new users, managing scalability for large datasets, and mitigating biases in recommendations. The motive of this study is to develop a robust movie recommendation system that enhances user satisfaction by offering content tailored to individual perceptions. This paper suggests a novel approach to develop an effective movie recommendation model using Bi-Directional Gated Recurrent Unit (Bi-GRU). The proposed Bi-GRU recommendation model is imposed on a Movie Lens dataset for movie preferences. Firstly, at the classification stage, Auto Regressive Integrated Moving Average and Convolutional Neural Networks (ARIMA-CNN) is employed to classify the movies with the assistance of temporal data. Furthermore, Bi-Directional Gated Recurrent Unit (Bi-GRU) is employed is incorporated for recommendation model as it can assist in summarization of the text and enhancement of the model this review from the audience signifies optimistic movie recommendation. The Bi-GRU model outperformed the existing method i.e. Unsupervised Boltzmann Machine-Based Time-Aware Recommendation System (UBMTR) by providing effective results in RSME (0.62), MAPE (0.38).