IMPLEMENTATION NAIVE BAYES ALGORITHM TO CLASSIFY HOAX NEWS
Daftar Isi:
- ABSTRACT News is information about current events that have a purpose to giving knowledge or warning the readers based pm events that already happen. In Indonesia, hoax news phenomena spread hate speech or strong emotion in their news content and social media to persuade the readers to get wrong information and do wrong action. This project is aimed to classify news and message into hoax or real news using Sentiment Analysis with Multinomial Naive Bayes Algorithm . To get more optimal result, there are two process in this project. First process is pre-processing data source using Document Frequency Thresholding and Term Frequency Inverse ( TF – IDF ) to gain information on the news. Second process is classify data using Multinomial Naive Bayes . The experiment showed the more its training data, the longer it takes to classify the test data. Experiment also showed the more specific the topic of data owned by training data then the performance of Multi-nominal Naive Bayes Algorithm will be better . Keyword: text classification, multinomial naive bayes, tf - idf, sentiment ana