🧠 Arabic Spell Checker – NLP-based Contextual Correction Tool

📋 Overview

This project presents an Arabic spell checker that leverages Natural Language Processing (NLP) techniques to provide intelligent, context-aware corrections. It aims to improve the accuracy and reliability of spell checking for Arabic, one of the most morphologically rich and complex languages.

🎯 Objectives

• Detect misspelled Arabic words.
• Suggest corrections based on context and linguistic patterns.
• Compare traditional edit-distance methods with probabilistic language models.

---

🧩 Methodology

🔹 Dataset

• Dictionary dataset: ~890,000 Arabic words for Levenshtein algorithm.
• Arabic corpus: 8,660 lines for the n-gram model.
• Both datasets were collected from open online linguistic resources.

🔹 Algorithms Used

1. Levenshtein Distance – for edit-distance-based correction.
2. N-gram Language Model – for context-based probability scoring.
3. Beam Search – for combining frequency-based scoring with contextual correction.
4. Tokenization, Stemming, POS Tagging – for text preprocessing and linguistic analysis.

---

🛠️ Tools & Technologies

• Language: Python
• Libraries: NLTK, NumPy, Tkinter
• NLP Techniques: Tokenization, Stemming, POS Tagging
• Interface: GUI built using Tkinter

---

🧮 Results

• The Levenshtein method effectively handled minor typographical errors (insertions, deletions, substitutions).
• The N-gram model achieved higher contextual accuracy and better prediction of common word sequences.
• Combining both approaches enhanced overall correction accuracy and flexibility.

---

🧰 How to Run

1. Open the UI folder.

2. Run the file:

   main.exe

   or

   python main.py

3. Enter Arabic text in the GUI to test the spell checker interactively.

---

👥 Group Members

• Jalila Moaddi – 1201611@student.birzeit.edu
• Maryan Kassis – 1200861@student.birzeit.edu
• Obada Hattab – 1171616@student.birzeit.edu

Department of Electrical and Computer Engineering – Birzeit University

---

🔮 Future Enhancements

• Expand the Arabic corpus for better contextual accuracy.
• Integrate modern BERT-based language models for semantic understanding.
• Optimize runtime efficiency and GUI usability.
• Develop a web-based interface for public testing.

---

📚 References

• Jurafsky, D., & Martin, J. H. (2024). N-gram Language Models. Stanford University.
• Saturn Cloud (2024). Stemming in Natural Language Processing.
• UBIAI (2023). NLP Techniques: Tokenization, POS Tagging and NER.
• ScienceDirect (2024). Levenshtein Distance Overview.