Sports News Summarization Using Ensebmle Learning

Abstract

This The process of producing a short version of documents by keeping the important information of documents is called text summarization. One of the ways to extract fundamental sentences is using optimization. Furthermore, optimizers try to specify the best weights for sentence features in order to select appropriate sentences. Although some text summarization researches use different optimizers but using ensemble learning can improve the performance of summarization systems. Because ensemble learning usually can cover more information than individually algorithm. In this paper, we show the effect of using ensemble learning for Persian sports news summarization. Therefore, we selected three well-known optimizers such as Genetic Algorithm (GA), Grey Wolf Optimizer (GWO), and Particle Swarm Optimization (PSO). In this study, we observe the performance of each of them and ensemble learning of the optimizers. We show that using ensemble learning cause better performance rather than using an optimizer in individual form. The evaluation metric that is used in this paper is F-measure. In addition, to evaluation approaches we gathered 10000 sports news from “Varzesh3” as our corpus.