<![CDATA[At this time, people can easily search and get information or news, both news through television and news from online media. The number of facilities that support the public to read the news causes the number of newsreaders in Indonesia increase too. However, many news articles found that the number of words and the use of words are less effective so it would be a waste of time when reading the entire contents of the news. From these problems, it takes a system that is able to summarize the content of the news in order for the news content to become dense. To summarize the content of the news, in this study used a method that is Maximum Marginal Relevance to produce a summary. In the method required several stages including, preprocessing, weighting TF-IDF, weighting cosine similarity and maximum marginal relevance method itself. This study was conducted by taking 30 samples of news article data with the theme COVID-19 from the website of online news provider kompas.com. Obtained the following test results, the best value regulator coefficient is α=0.5 with precission result = 0.684333, recall = 0.772 and f-measure = 0.7. While based on the number of words, the number of words translated to 300 produces the best f-measure value with a value of 0.726923. As well as being tested systems with and without stemming and the result the system using stemming produces a better summary than the system without stemming.]]>
Copyrights © 2021
