An Intelligent System for Automatic Classification of Academic Titles Using AI and NLP

Abstract

This project aims to propose a system for classifying book and thesis titles using Artificial Intelligence (AI) and Natural Language Processing (NLP). It employs the DeBERTaV3 model, trained on a multi-category academic dataset, and supports multilingual input through OCR, language detection (langdetect), and automatic translation (googletrans). The system achieved high accuracy in classifying titles into fields such as Economic Sciences, Law, and Technical Sciences. Developed in Python, it integrates libraries such as Transformers, Pandas, PyTorch, and PySide6, with results stored in an Excel database.