Sentiment Analysis App

Perform sentiment analysis using the traditional bag of words technique with TF-IDF weighting.

Powered by Scikit-learn and Streamlit.

Running Locally

NOTE: Requires Python 3.10 and above.

1. Download the code, and create a virtual environment:

   git clone https://github.com/Tim-Abwao/sentiment-analysis-app.git
   cd sentiment-analysis-app
   python3 -m venv venv
   source venv/bin/activate

2. Install dependencies:

   pip install -U pip
   pip install -r requirements.txt

3. Launch the app:

   streamlit run streamlit_app.py

Datasets

The datasets used here are subsets of the Amazon Customer Reviews Dataset. See datasets for more details.

NOTE: The Bag of Words model has known limitations (high dimensionality, poor comprehension of context, etc). To keep the app within Streamlit Cloud's resource limits, the samples used are rather small.