Deep learning-based sentiment analysis of Netflix reviews using LSTM and BI-LSTM

The rapid growth of Over-The-Top (OTT) streaming platforms has led to a significant increase in user-generated content in the form of online reviews. These reviews provide valuable insights into user satisfaction, preferences, and overall platform performance. In recent years, online streaming platforms like Netflix have generated a huge number of user reviews, which reflect customer opinions and experiences. Analysing these reviews manually is difficult because of the large volume of data. To solve this problem, this study focuses on using deep learning techniques to automatically identify the sentiment of Netflix reviews.
A dataset containing more than 146,000 reviews was used for this research. The text data was first cleaned and processed using Natural Language Processing techniques such as tokenization, stop word removal, and padding. The reviews were then classified into two categories: positive and negative.
This study uses two deep learning models, LSTM and Bidirectional LSTM (Bi-LSTM), to understand the sequence and context of words in the reviews. While the LSTM model captures long-term dependencies, the Bi-LSTM model improves performance by analysing the text in both forward and backward directions. The results show that the Bi-LSTM model performs better, achieving an accuracy of around 89.8%.
Overall, this work demonstrates that deep learning models are effective in analysing large-scale textual data and can help platforms like Netflix better understand user feedback and improve their services.