Performance Evaluation of Nature-Inspired Metaheuristic Approaches for Single Document Text Summarization

Abstract

In today era, day by day huge amount of data is collected on internet. The reading of text document or retrieving important information are time consuming process, so there is need for introducing effective text summarization technique. Text summarization, is the process of retrieving key information from lengthy document, its plays an essential role in information retrieval and content extraction. The paper we presented a comprehensive examination of nature-inspired metaheuristic algorithms, such as firefly, Cuckoo Search(CS) and Particle Swarm Optimization (PSO) to improve text summarization with an emphasis on single document datasets such as DUC-2001 and DUC-2002. The measurement of generated text summaries quality, generated summaries of datasets are compared with existing golden summaries and evaluated using ROUGE score. Our results show that nature-inspired metaheuristic-based approaches show potential for enhancing text summary of individual documents, metaheuristics methods improve summarizing effectiveness while offering a fresh viewpoint on how to handle the process within the confines of a single document dataset.