Penerapan Rocchio Classification dalam Klasifikasi Testimoni Pelanggan Taksi

Abstract

Testimonials are expressions of users of products or services in the form of satisfaction or laughter. When testimonials contain satisfaction, it can be the attraction of a product or service in getting consumers. Conversely, if the testimonial is in the form of disappointment then it becomes an evaluation material and a big task for the owner or management of the company to improve the services that have been provided. Testimonials we can get from social media such as twitter, instagram, etc. Testimonial classification by reading and classifying manually is not effective if testimonials in large numbers. Therefore, in terms of assisting the evaluation of services that have been provided, testimonial classification is carried out by applying the Rocchio Classification algorithm. Testimonials used in this study are testimonials of taxi customers obtained from twitter. The data is processed in the preprocessing stage, tf-idf, classification process. The results showed that Rocchio has a fairly high accuracy of 80%, an average precession of 83.40%, an average recall of 63.28%, an average f-measure of 65.16%.