PENERAPAN ALGORITMA DECISION TREE UNTUK ANALISIS SENTIMEN ULASAN APLIKASI VIDIO DI SITUS GOOGLE PLAY

Video streaming-based applications are becoming the most widely used type of application in the world. Video On Demand is an interactive system that allows us to choose the video content to watch. Vidio is an online portal or video streaming website established in 2014. The website allows users to watch and enjoy various videos and other services. However, based on reviews on Google Play, Vidio received an average rating of only 3.7 from more than 623,000 total reviews. This is what prompted this research. The data collected is as many as 1000 in the period of February 2, 2023 – February 19, 2023. The data is classified into positive and negative sentiment using the Decision Tree algorithm. Based on 3 data sharing scenarios, the greatest accuracy was obtained from data sharing of 80% of training data and 20% of test data, which was 97.3%. While in the 70:30 data sharing scenario, the accuracy is 96.8%, and the 90:10 data sharing is 96.8% accuracy. From the accuracy that has been obtained, for the evaluation of model testing, this study uses the Confusion Matrix. In order for predictions from the trained model to be available to others, this study conducts model deployment using Streamlit.

Keywords: Video Streaming, Vidio, Decision Tree, Confusion Matrix, Streamlit