OPINION MINING PADA KOMENTAR YOUTUBE MENGGUNAKAN ALGORITMA K-MEANS

Public opinion on YouTube video comments is a potential data source to find out public interest in a certain topic. Opinion Mining is a technique for processing and obtaining information from an opinion sentence. This research developed an Opinion Mining software using K-Means algorithm that could classify comments into three groups, those are positive, negative, and neutral. This study using word weighting technique and some corpus in preprocessing and weighting sentence. The results showed that the software was able to classify comments with 91% accuracy of the test results of 500 comments. This research can be developed better if there is an update corpus process for every sentence in the implementation.