期刊名称：Karbala International Journal of Modern Science
印刷版ISSN：2405-609X
电子版ISSN：2405-609X
出版年度：2017
卷号：3
期号：3
页码：119-130
DOI：10.1016/j.kijoms.2017.06.001
语种：English
出版社：Elsevier
摘要：Abstract With the exponential growth of information in World Wide Web, extracting relevant information from huge amount of data has become a critical task. Text summarization has been appeared as one of the solution to such problem. As the main objective is to retrieve a condensed document that pertain the original information, so it can be considered as an optimization problem. In this paper, a comparative analysis of few meta-heuristic approaches such as Cuckoo Search (CS), Cat Swarm Optimization (CSO), Particle Swarm Optimization (PSO), Harmony Search (HS), and Differential Evolution (DE) algorithm is presented for single document summarization problem. The performance of all these algorithms are compared in terms of different evaluation metrics such as F score, true positive rate and positive predicate value to validate summary relevancy and non-redundancy over traditional and standard Document Understanding Conference (DUC) datasets.
