This application performs sentiment analysis using a Deep CNN model built with TensorFlow. It provides a GUI where users can enter text and see whether the sentiment is positive or negative.
The model was trained on the Sentiment140 dataset from Kaggle. Training statistics:
loss: 0.1550 - accuracy: 0.9379
To run this application, you need to have the following packages installed:
- TensorFlow
- NumPy
You can install them using pip:
pip install tensorflow numpy
After ensuring you have the required packages, simply run the provided main.py script. The GUI will show up, allowing you to enter a text and analyze its sentiment.
- The main window is titled "Sentimental Analysis" and provides an input field where users can enter their text.
- After entering the text, click on the "Submit" button to get the sentiment result.
- A message box will pop up displaying whether the entered text has a positive or negative sentiment, along with a confidence score.
The DeepCNN class in the code defines the architecture of the Convolutional Neural Network. It utilizes embeddings followed by multiple convolutional layers with varying kernel sizes to capture bi-grams, tri-grams, and quad-grams from the input text. The final layers include dense layers and a dropout layer for regularization.
Application and Model created by Ali Mohammad and Dawi Alotaibi.