Multi-document Extraction Text Summarization via Whale Swarm Optimization

In today's world, the amount of knowledge available on the web is increasing rapidly, thus making automatic text summarization an increasingly important area of research. Multiple-document summarization presents more challenges than single-document summarization does. This paper proposes a new 'whale optimization-based text summarization' for multiple documents. It uses a single objective function that uses three features: closeness to the topic, cohesion, and degree of readability. The experiments used four different benchmark datasets from the DUC-2002, DUC-2004, DUC-2006 and DUC-2007. The recall-oriented understudy for gisting evaluation (ROUGE) was used to evaluate the summarizer results. The results show that the existing summarizers in the literature were improved significantly in terms of the average ROUGE-F score by the proposed summarizer from the experimental results.