Özet:

The amount of data available in the electronic environment is increasing day by day with the development of technology. It becomes tough and time consuming for the users to access the information they desire within this increasing amount of data. Automatic text summarization systems have been developed to reach the desired information within texts in a shorter time than that of manual text summarization. In this paper, a new extractive text summarization model is proposed. In the proposed model, the inclusion of sentences of a given text in the summary is decided based on a classification approach. Also, the effectiveness of widely used features for automatic text summarization in the Turkish language is evaluated using sequential feature selection methods. The evaluations were carried out specifically for Turkish texts in the categories of economy, art, and sports. The experimental work justified the performance of the proposed text summarization method and revealed how effective the features are.

Anahtar Kelimeler: