Single document text summarization addressed with a cat swarm optimization approach.

Area of Science:

• Natural Language Processing
• Artificial Intelligence
• Computational Linguistics

Background:

• The proliferation of online information necessitates efficient methods for extracting key data.
• Automatic text summarization aims to condense large documents into concise summaries.
• Existing methods often struggle with content coverage, redundancy, and readability.

Purpose of the Study:

• To propose a novel Cat Swarm Optimization (CSO) algorithm for single-document extractive summarization.
• To enhance summary quality in terms of content coverage, informativeness, anti-redundancy, and readability.
• To evaluate the proposed CSO-based summarization system against state-of-the-art methods.

Main Methods:

• Pre-processing of input documents.
• Initialization of a cat population with binary vectors representing sentence selections.
• Formulation of an objective function based on sentence quality metrics.
• Iterative optimization using CSO's seeking/tracing modes and a Best Cat Memory Pool (BCMP).

Main Results:

• Achieved approximately 25% and 5% improvement on ROUGE-1 and ROUGE-2 scores, respectively, over existing methods on DUC-2001 and DUC-2002 datasets.
• Demonstrated superior performance in content coverage, informativeness, and anti-redundancy.
• Evaluated summaries for readability, conciseness, relevance, and processing time, showing significant advantages.

Conclusions:

• The proposed Cat Swarm Optimization (CSO) algorithm offers a superior approach to automatic text summarization.
• The system generates high-quality, readable, and concise summaries efficiently.
• Statistical tests confirm the significance and effectiveness of the CSO-based summarization method.