Tota Simatupang
Customer reviews are an essential reference to measure product performance in many aspects. Social media is suitable as a customer review data source because it has many online volunteers who write their reviews. To analyze online customer reviews, artificial intelligence is needed. Many studies have applied the artificial intelligence approach, but few studies focus on the courier service industry in Indonesia. The courier service industry has proliferated in the last few years. Thus, the companies in this industry need to reach a broader market and understand their customers’ preferences promptly. Therefore, in this research, text mining and sentiment analysis methods were applied to analyze the customers’ preference. In this research we used PT. Pos Indonesia courier services as a case study. Four hundred fifty customer reviews were collected from REST API Twitter in Bahasa Indonesia for three products of PT. Pos Indonesia. The sentiment analysis approach used in this research is the supervised machine learning and Support Vector Machine (SVM) algorithm. We built the model using R Studio Application. The model classifies the data into three classes: Negative, Positive, and Neutral. The model we built has an accuracy of 97%. The values of five performance measures (recall, specificity, precision, F1 score) are more than 90%. Thus, we conclude that the model is good enough at conducting sentiment analysis for courier service review data of PT. Pos Indonesia. Then, we design the application of the model to make it easier for the user to see the result of the product evaluation.
Penerapan Karya Tulis
To analyze online customer reviews, artificial intelligence is needed.