<![CDATA[In the digital age, film reviews on social media platforms and review sites have become an important source of information about public opinion. However, manually processing a large number of film reviews is very difficult and time-consuming. Therefore, this research focuses on sentiment analysis of film reviews using natural language processing (NLP) techniques and BiLSTM, an extension of recurrent neural network (RNN) models. Using a data set of 50,000 film reviews from IMDB, this study ran four different scenarios to train the BiLSTM model. Results showed that Model 3, with Embedding Dimension 64 and Batch Size 32, gave the best performance with an accuracy of 0.8727 which is a good accuracy as it means that 43,635 out of 50,000 reviews were predicted correctly. The research conclusion highlights the effectiveness of BiLSTM in analyzing sentiment on movie reviews, particularly in overcoming the challenges of long text.]]>
Copyrights © 2024
