A lightweight Natural Language Processing (NLP) pipeline that performs rule-based sentiment analysis on textual review data. This project efficiently ingests bulk text data, processes it through standard NLP lexicons, and generates visual distributions of sentiment.
This script utilizes NLTK's VADER (Valence Aware Dictionary and sEntiment Reasoner) to calculate compound polarity scores for text entries. The scores are classified into Positive, Neutral, or Negative categories. The data processing utilizes functional programming paradigms via Pandas for optimized execution across large datasets.
The data used for this project is not hosted in this repository due to file size constraints. You can download the original dataset here: https://shorturl.at/hOs3g
Note: Once downloaded, place the dataset in the root directory and ensure it is named
reviews.csv.
- Python 3.x
- NLTK (VADER) - Core NLP sentiment lexicon.
- Pandas - Tabular data manipulation and functional application (
progress_apply). - Seaborn & Matplotlib - Data distribution visualization.
- TQDM - Terminal progress tracking for bulk operations.
- Clone the repository:
git clone [https://github.com/CLOVEOS/sentimentAnalysis.git](https://github.com/CLOVEOS/sentimentAnalysis.git)
cd sentimentAnalysis