🔍 Job Recommender NLP: AI-Powered Job & Talent Matching with Sentence-BERT

📘 Overview

This project builds an AI-powered job and resume recommendation engine using Sentence-BERT embeddings to match resumes with job postings — and vice versa. It was initially developed as a demo on Kaggle to showcase how semantic similarity can improve recruitment and job searching over traditional keyword-based systems.

The system has now been enhanced with a Streamlit web interface for interactive use and deployment.

It processes a subset of the LinkedIn Job Postings Dataset (2023-24) and the Resume Dataset, recommending the top 5 matches in both directions:

• Resumes per job posting (for recruiters)
• Job postings per resume (for job seekers)

---

🔗 Links

• 📘 Kaggle Notebook: Job Recommender NLP
• 💻 GitHub Repository: Job-Recommender-NLP
• 📄 LinkedIn Job Postings Dataset: View Dataset
• 📄 Resume Dataset: View Dataset

---

🔑 Key Features

🧠 Bi-Directional Semantic Recommendation Engine

• Matches 1,000 job postings with 1,000 resumes using semantic similarity
• Recommends:
  • 🔍 Top 5 jobs for an uploaded resume (job-seeker perspective)
  • 🧑‍💼 Top 5 resumes for an uploaded job description (recruiter perspective)

⚙️ Implementation Details

• Built with Sentence-BERT (all-MiniLM-L6-v2) for embeddings
• Features:
  • Text preprocessing using spaCy (lemmatization, skill/domain extraction)
  • Cosine similarity computation
  • Interactive frontend using Streamlit
  • Auto-generation of working files if missing

🌟 New Features

• 🎨 Visually Appealing UI: Custom-styled Streamlit interface
• 📁 File Uploads: Supports .txt, .pdf, .docx for resumes and job descriptions
• ⚙️ Dynamic Setup: Automatically creates the working directory if absent

---

📁 Repository Contents

• app.py: Streamlit app for interactive resume/job matching
• job_recommender_nlp.py: Backend logic (data processing, embeddings, recommendation)
• job-recommender-nlp.ipynb: Jupyter Notebook version
• datasets/: Includes postings.csv and Resume.csv

---

🚀 How to Run

✅ Prerequisites

• Python 3.10+

Install dependencies:

pip install numpy pandas spacy sentence-transformers scikit-learn PyPDF2 python-docx streamlit

Install the spaCy model:

python -m spacy download en_core_web_sm

---

🛠️ Local Setup

1. Clone the Repository

git clone https://github.com/MuhammadAliAsgher/Job-Recommender-NLP

cd Job-Recommender-NLP

2. Prepare Datasets

• Datasets (postings.csv, Resume.csv) are inside the datasets/ directory

• Managed using Git LFS — large files auto-downloaded when you clone the repo

Run the App

streamlit run app.py

---

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

🙋‍♂️ Contact

Created by Muhammad Ali Asghar
📧 Connect on LinkedIn
🌐 Github/MuhammadAliAsgher