Gojek Customer Sentiment Analysis

This project uses data science techniques to analyze user sentiment towards the Gojek application in Indonesia by scraping user reviews from the Google Play Store, preprocessing the text data (cleaning, case folding, tokenization, stop word removal, slang word replacement, filtering, and stemming), and performing sentiment analysis using both a lexicon-based approach and machine learning. The project involves extracting features from the text using the TF-IDF technique, addressing class imbalance with SMOTE, and training a Support Vector Machine (SVM) model to classify customer reviews as positive or negative. This process identifies the overall sentiment (positive, negative, or neutral), visualizes the sentiment distribution, explores the most frequent words associated with each sentiment category, and ultimately provides insights into customer satisfaction and areas for potential improvement for Gojek.