Automatic Summarization of Scientific Articles

Authors:

Sanjana G, Aishwarya P, Aparna K, Sujoy Sarkar

Abstract—The rapid growth of digital libraries and online research repositories has resulted in a massive increase in the number of scientific articles available to researchers. While this growth enables knowledge sharing, it also creates challenges for researchers who need to quickly identify relevant information from large research papers. Manual reading of lengthy scientific documents is time-consuming and inefficient. Therefore, auto- mated summarization techniques are required to generate concise summaries that capture the essential information of research articles.

This project presents an automated system for summarizing scientific research papers using Natural Language Processing (NLP) and transformer-based models. The system extracts textual content from uploaded PDF files, processes the extracted text using preprocessing techniques, and generates meaningful summaries using machine learning and deep learning models. The framework also integrates an AI-based chatbot that allows users to ask questions related to the uploaded research paper and receive contextual answers.

The proposed system combines traditional text preprocessing, NLP-based summarization algorithms, and transformer models to produce accurate and informative summaries. The system architecture includes modules for PDF text extraction, data preprocessing, summarization, chatbot interaction, and result visualization. Experimental results demonstrate that transformer-based summarization models outperform traditional extractive methods by generating more coherent and contextually meaningful summaries.

This research contributes to the development of intelligent document processing systems that assist researchers in quickly understanding scientific literature. The proposed system improves efficiency in academic research by reducing the time required to read and analyze large research papers.

Index Terms—Text Summarization, Natural Language Processing, Scientific Articles, Transformer Models, PDF Processing, AI Chatbot