Sentiment Analysis Using K-NN Algorithm Through Random Search Cross Validation Approach

This research discusses the application of the KNearest Neighbor (K-NN) algorithm in sentiment analysis, specifically in classifying positive, negative, or neutral sentiments from ChatGPT user opinions collected from user reviews on the Play Store. This research focuses on optimizing the key parameters of the K-NN algorithm using the Random Search cross-validation (RSCV) method. The evaluation results showed a good performance improvement after applying RSCV. Previously, the model showed unsatisfactory performance but improved significantly after optimization. Overall, RSCV successfully improved the average performance of the model for sentiment analysis.