This project is an attempt to create a Customer Review Intelligence Platform that helps us analyze large volumes of customer reviews to extract valuable insights. It uses Sentiment Analysis, Topic Modeling, and Trend Analysis to summarize customer opinions, detect trends, and visualize actionable insights through an interactive dashboard.
- 1.Data Pipeline
- 2.Advanced NLP Processing
- 2.1 Semantic Analysis
- 2.2 Topic Modelling
- 2.3 Feature Engineering
- 2.4 Trend Detection
- 3.ML Models -Pretrained models like distilBERT
- 4.Interactive Dashboard
This project uses the "McAuley-Lab/Amazon-Reviews-2023" dataset from Hugging Face. The dataset was downloaded using the datasets library and stored locally as a CSV file for efficient loading and processing. No need to manually download the dataset. link: https://huggingface.co/datasets/McAuley-Lab/Amazon-Reviews-2023
I have created separate .py notebooks for different implementations that are independently reusable and scalable for future changes etc. The following define how each one functions:
- main.py: connects everything and is responsible for workflow
- data_loader.py: dataset loading
- sentiment_analysis.py: applies sentiment analysis
- rend_analysis.py: generates trend insights
- dashboard.py: launches the interactive dashboard