Summarizing Indonesian text automatically by using sentence scoring and decision tree

Abstract

Text summarization is a process of compressing a text from the source to be a shorter version, but the version still contains the main information there. By reading the summary, the readers might be easy and fast to understand the contents instead of reading all the text. Because of that, it needs a method to understand, clarify, and present the whole information needed clearly and succinctly in the summary. So, it allows the readers save the time and energy. This research combining sentence scoring and decision tree method for automatic text summarization in Indonesian language. It uses the decision tree algorithm to choose which of sentences will be selected in summarization system. To produce the rules for decision tree, it uses 50 news texts as the training data. The produced-model from the training stage will be implemented for sentence selection process to the summarization system. The result shows the highest f-measure score is 0, 80 and the average is 0, 58. Based on this, it concludes that the result of document summarization using sentence scoring and decision tree shows a better accuracy score for news text document.