ANALISIS SENTIMEN PADA OPINI PENGGUNA APLIKASI QASIR MENGGUNAKAN SUPPORT VECTOR MACHINE DAN RANDOM FOREST

Abstract

Qasir is an Android-based Point-Of-Sale (POS) application that can be accessed for free on the Google Playstore. With so many POS applications available, users will be more selective in choosing the application to use. One aspect that can influence the decision to choose an application is the opinion on the application. Opinion is information obtained after using the application that can contain criticism or suggestions. So based on this the user can conclude how other users use the application. Besides being useful for users, opinions if processed properly will produce information that can be used for evaluation for the development team. To analyze and find relationships between owned data, you can use Data Mining. This research will use the Support Vector Machine and Random Forest methods, but each method has its advantages and disadvantages so that the accuracy of the two methods will be compared. The results obtained are that the Support Vector Machine has the highest accuracy value with 80.63% while the Random Forest is 80.21%.