KLASIFIKASI TWEET SPAM DAN VALID MENGGUNAKAN SELEKSI FITUR CHI SQUARE DAN ALGORITMA NAÏVE BAYES CLASSIFIER PADA TWEET BERBAHASA INDONESIA

RAKHMAN, DERTA ISYAJORA, Aina Musdholifah

2016 | Skripsi | FMIPA

Twitter users have no limitation to send any tweet that contain diversified information, including spam. JalananYogya is one of the application that use that feature. JalananYogya is crowdsourcing platform that leverages community participation to report damaged roads in Yogyakarta using Twitter. To maintain JalananYogya’s data integrity, system that can perform filtering on valid report and spam report is required As a first step in making spam filtering system, potential classification algorithms will be tested. The focus of this research was to evaluate the performance of Naïve Bayes Classifier algorithm to classify tweet combined with Chi Square feature selection to improve performance. The method used by this reasearch is Multinomial Naive Bayes Classifier and Bernoulli Naïve Bayes Classifier. Based on the tests performed, 95 % is the best accuracy produced by systems that have been built on this research. That system using Multinomial Naïve Bayes Classifier methods combined with Chi Square feature selection

Kata Kunci : Tweet, JalananYogya, Tweet, Spam, Naïve Bayes Classifier, Chi Square