<![CDATA[JNE and J&T Express are two of the largest and most popular courier companies in Indonesia, leading to various public opinions regarding the quality of their services. This research employs a lexicon-based method using the InSet dictionary, a simple scientific approach where the system calculates the weight of words and classifies them as positive, negative, or neutral sentiments. The analysis process begins with data collection of reviews using scraping techniques, followed by text processing including cleaning, case folding, normalization, tokenization, stemming, and stopword removal. Out of 3,565 reviews for JNE and 3,967 reviews for J&T, the sentiment analysis indicates that the majority of the public holds negative opinions towards the services of both courier companies. The analysis accuracy reaches 82% for JNE data, with a precision value of 95% for negative sentiment, 54% for positive sentiment, and 7% for neutral sentiment. The sensitivity values are 83% for negative sentiment, 82% for positive sentiment, and 15% for neutral sentiment. Data for J&T shows an accuracy of 78%, with a precision value of 97% for negative sentiment, 28% for positive sentiment, and 4% for neutral sentiment. Sensitivity values are 80% for negative sentiment, 82% for positive sentiment, and 4% for neutral sentiment.]]>
Copyrights © 2025
