Meme Opinion Categorization by Using Optical Character Recognition (OCR) and Naïve Bayes Algorithm

Generally, a meme is an image which is produced by the society which is used to comment a certain event, followed with a particular template from the decent online images. The distribution of meme becomes the phenomenon and very popular in the last few years. The problem is, some of these memes contain negative contents to harm others. One type of meme image that trends in social media is aimed at government, either to support the performance of the government or to insinuate and dislike of a government. Therefore, Political view by the citizen can be identified by viral memes on the internet. The aim of this research is classifying the types of a meme by applying image processing and OCR Tesseract which are combined with Naïve Bayes Algorithm. OCR Tesseract is required to recognize text in an image, meanwhile Naive Bayes algorithm which is used to find the highest probability to classify the testing dataset into the correct category. This research uses a meme as the dataset. The result is meme which is successfully classified. The accuracy depends on the OCR result which utilizes tesseract engine.