SAMIKSHA – Sentiment Analysis of E-Consultation Comments

📌 Problem Statement

ID: 25035
Title: Sentiment analysis of comments received through E-consultation module
Theme: Miscellaneous
Team ID: VJITIH-15
Team Name: SAMIKSHA

---

💡 Proposed Solution

We aim to build a modular, multi-tier web application for analyzing feedback comments.
The key features include:

• Process comments individually or in bulk (via file uploads).
• Classify comments into Positive, Negative, or Neutral sentiment.
• Extract keywords, topics, and entities from feedback.
• Summarize long comments into concise summaries.
• Provide a web-based dashboard for real-time visualization.

---

⚙️ Technical Approach

• Backend: Python, Flask API
• AI/ML: Scikit-learn, NLTK, Transformers (NLP models)
• Frontend: React.js, TailwindCSS
• Visualization: Matplotlib, WordCloud
• Deployment: Vercel

---

✅ Feasibility & Challenges

1. Misunderstood comments → Sarcasm/complex language can confuse AI.
2. Badly written comments → Poor input reduces accuracy.
3. AI dependency → May miss important insights.
4. Setup time → Requires effort to train models initially.

---

🌍 Impact & Benefits

• ⏳ Saves Time: Handles thousands of comments quickly.
• ⚖️ Fair & Consistent: AI removes personal bias.
• 🔎 Easy Insights: Word clouds highlight key issues.
• ✂️ Quick Summaries: Long inputs → concise insights.
• 📊 Scalable: Works efficiently even with huge volumes of data.

---

📊 Dataset & References

• Dataset: Twitter Entity Sentiment Analysis – Kaggle
• Twitter API Docs: X API Docs
• Scikit-learn Docs: v1.7 Release
• NLTK Installation: NLTK Docs
• Research Paper: Sentiment Analysis in Social Media Using Deep Learning

---

🚀 How to Run

Backend (Flask)

cd backend
pip install -r requirements.txt
python app.py

Frontend (React)

cd frontend
npm install
npm rundev

---

👥 Team – SAMIKSHA

• Team ID: VJITIH-15
• Contribution: AI/ML, Web Development, Data Analysis