Comparative Analysis of Machine Learning and Deep Learning Techniques for Extractive Text Summarization

Abstract

In the age of digital content overload, extractive text summarization serves as a critical tool for information compression and retrieval. While both machine learning (ML) and deep learning (DL) techniques have shown promise in this domain, there remains a lack of comprehensive comparisons across diverse models and datasets. This study conducts a comparative analysis of multiple ML and DL methods for extractive summarization using two benchmark datasets: BBC News articles and summarized IMDB reviews. Traditional ML models such as k -Nearest Neighbors, Decision Trees, and Random Forests are evaluated alongside advanced DL architectures including BiLSTM with Attention and BERTSum. All models follow a uniform preprocessing pipeline, and their performance is measured using ROUGE metrics. Experimental results show that while deep learning models generally outperform traditional methods in terms of precision and contextual understanding, certain ML models remain competitive due to their efficiency and interpretability. This work offers practical insights into the trade-offs between ML and DL approaches and guides future research in selecting appropriate models based on summarization needs.