KLASIFIKASI SENTIMEN TWITTER TERHADAP TRANSPORTASI OJEK ONLINE MENGGUNAKAN METODE LEARNING VEKTOR QUANTIZATION 2 (LVQ 2)
Daftar Isi:
- Twitter social media is widely used by the community to convey sentiments towards ojek online transportation. Data from community sentiments towards ojek online transportation can be transformed into information through sentiment analysis. Therefore, this study aims to classify sentiment data into 2 sentiment classes, namely positive sentiment, and negative sentiment. The calcification method used in this study is Learning Vector Quantization 2 (LVQ2). The results of the system show that the LVQ2 algorithm produces an optimal accuracy of 94.37% with a combination of learning rate value parameters of 0.1, widow value 0.2, the learning rate multiplier value is 0.6, the maximum number of iterations is 6 iterations, the percentage of the training data is 90%, with the training data 455, and the testing data is 55 data, and the minimum alpha is 0.01.