📚 AI Resume Screening App

A Streamlit-based NLP-powered resume screening tool that classifies resumes into job categories and ranks them by relevance. It also compares resumes against job descriptions using cosine similarity.

---

🚀 Features

• 📄 Extracts and cleans resume text from .pdf, .docx, and .txt
• 🧠 Predicts job categories using a trained ML model
• 📊 Ranks multiple resumes by prediction confidence
• 📝 Compares resume content with pasted job descriptions
• 💬 Collects user feedback

---

🔍 Technologies Used

• Python
• Streamlit
• Scikit-learn
• NLTK
• Plotly
• NumPy, pandas, re, PyPDF2, python-docx

---

📦 Folder Structure

resume-screening-app/ │ ├── app.py # Streamlit app interface ├── clf.pkl # Trained classifier (multiclass) ├── tfidf.pkl # TF-IDF vectorizer ├── encoder.pkl # Label encoder for categories ├── requirements.txt # All dependencies ├── UpdatedResumeDataSet.xls # Resume dataset used for training ├── README.md # You're reading this!

---

---

✅ Requirements

pip install -r requirements.txt

---

🚀 Run the App

streamlit run app.py

---

##🧠 Model Training (optional) The classifier was trained using Scikit-learn on a labeled resume dataset with a TF-IDF vectorizer and a Multinomial Naive Bayes classifier. To reproduce the training pipeline, see the notebook: Resume Screening with Python.ipynb

---

📌 Example Predictions

• ✅ Filename: resume1.pdf

• ✅ Predicted Category: Data Science

• ✅ Confidence: 92.5%

• ✅ JD Similarity Score: 88.7%

---

🌟 Contribution

• Pull requests and enhancement ideas are welcome!
• Feel free to fork the repo and add new features or suggest improvements.

---

👩‍💻 Author

Rishita Makkar 📧 Email: rishita.m@example.com 🌐 LinkedIn (Update with your profile)

🏷️ Tags

#NLP #Streamlit #MachineLearning #ResumeClassifier #AIRecruitment #PortfolioProject

---

✅ How to Use This

1. Save this file as README.md in the root directory of your project.
2. Push it to GitHub — it will render beautifully and make your repo more attractive.

Let me know if you want me to generate the matching requirements.txt or zip the full folder for upload.

---

✅ 2. requirements.txt

Generate this automatically (recommended):

pip freeze > requirements.txt