⚡ EmotionLens – ML Optimized Version (v2)

(Training & Model Optimization)

(Testing & Prediction Pipeline)

🔍 Overview

This is the upgraded version of my EmotionLens project — featuring machine learning–based sentiment analysis, optimized NLP pipelines, and deeper emotional understanding of Amazon product reviews.

It builds upon my earlier project EmotionLens (EDA Foundation), which focused on text cleaning, EDA, and rule-based sentiment analysis (VADER).

---

🚀 Key Highlights

• Implemented ML models (Logistic Regression, SVM, BERT)
• Added TF-IDF vectorization for text feature extraction
• Integrated emotion detection using HuggingFace Transformer
• Optimized preprocessing & visualization pipeline
• Evaluated using accuracy, F1-score, and confusion matrix

---

🧠 Workflow

1. Data Loading & Cleaning
2. Exploratory Data Analysis (EDA)
3. Feature Engineering (TF-IDF, Word2Vec)
4. Model Training & Evaluation
5. Emotion Detection using Transformers
6. Visualization & Insight Extraction

---

📂 Files

File Name	Description
Emotion_lens5.ipynb	Main ML pipeline for sentiment & emotion detection
Emotion_Test.ipynb	Model evaluation and testing notebook
requirements.txt	Required Python dependencies

---

📸 Visual Highlights

Emotion Distribution	Model Evaluation

---

▶️ Usage

Run on Google Colab:
Open EmotionLens ML Optimized → click Runtime → Run All

Run locally:

git clone https://github.com/Krish5986/EmotionLens-ML-Optimized.git
cd EmotionLens-ML-Optimized
pip install -r requirements.txt
jupyter notebook Emotion_lens5.ipynb