Sentiment analysis using different types of Bidirectional Recurrent Neural Networks on Amazon reviews dataset. The results are confronted with two baseline models which are an SVM and a RF model.
Jupyter Notebook
DLA-SentimentAnalysis
How to run using precomputed models
Open a terminal in the desiderate path and download the project folder using the command.
With the same terminal enter the project folder and install the requirements using this commands.
cd DLA-SentimentAnalysis
pip3 install -r requirements.txt
Download and extract the dataset in the project folder DLA-SentimentAnalysis/dataset from here
You should now have a directory called archive in the dataset folder, pick the file "train.ft.txt.bz2" and move it one level down in the "DLA-SentimentAnalysisDataset" folder.
You are ready to train, for starting the training and evaluation paste the following command in the terminal you opened in the step 0 (It must be located in the "DLA-SentimentAnalysis" folder).