Automatic Text Summarization Menggunakan Metode Graph dan Metode Ant Colony Optimization

Abstract

Many documents were distributed on the Internet every day. Not all of the text in a document is important information. To seek information from documents that are widely used methods Information Retreival (IR). Research in the field of Information Retreival (IR) has been made since the 1950s. Document retrieval system is a system that can search for documents by keyword, but the retrieval of the document will encounter problems if they have to seek specific information from the document a polynomial with the number of constituent documents of text that amount is not small. Automatic Text Summarization combines two methods of graph method and the method of ant colony optimization and use features four sentences. Four features the phrase used, namely: the similarity between sentences, sentences that resembles the title of the document TF-ISF and TF-IDF. Tests performed include: testing the summary results of the system with the summary results of the manual to get the 76.3%, the test results summary system with the summary results Autosummary tools in Microsoft Word to get the 68.15% and the test results summary system with the summary results of the expert that high school teacher, getting the 78.43 %. Results compression rate to get the system summary 78.2%.