SOCCER OPINION ANALYSIS OF TWITTER MEDIA USING BACK PROPAGATION METHOD

Twitter is a social networking tool that allows users to send and read text-based messages, commonly referred to as tweets. Information and news that is done by general users right now is no longer exclusive, it can only be done by big news publishers, but it can be done by everyone. Twitter has many opinions, not only positive or neutral opinions but also negative ones. In this research, each user's tweet will be categorized into positive and negative sentiment by looking at the contents of the tweet measured using TF-IDF weighting and Backpropagation Artificial Neural Network (ANN) classification method. Before using with TF-IDF, the contents of the tweet are done in the process of preprocessing data using tokenisais, cleansing, filtering, and Steming. To facilitate the work process on the system that was created, the data used is tweet in Indonesian. This research is a sentiment towards Twitter data. After doing research on Twitter data, it can be seen the results of trials of several scenarios with the amount of data 1000 and 1500 Tweet data from the classification of Backpropagation Neural Networks with the best 1000 data scenarios in table 4.12 with an accuracy of 50.58%, a precision level of 100.00 %, Recall 50.44%, and F-measure 67.06% so when changes in learnigrate or testing parameters inputted by the user affect changes in the accuracy of the system.