#TITLE_ALTERNATIVE#

• Main Author: TALITA (NIM : 13413012), NADYA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/29505
• Institution: Institut Teknologi Bandung

<p align="justify">Indonesia's recent surging growth in e-commerce is also affecting its related industries, such as courier service and logistic companies. PT. Pos Indonesia, as one of the main player of the courier service industry has experienced an increasing competition specifically from JNE Indonesia as the e-commerce market opportunity growing. The number of PT. Pos Indonesia’s customer less the JNE’s. To increase customers, information regarding user's need and preference is needed to enhance customer's satisfaction. Customer's twitter data is one of many data that can be used. Twitter enormous data containing customer's preference towards courier service in a large amount and the data has potential to be analyzed. The data's large amount makes it hard to be analyzed manually, so that text mining is used to analyze the data automatically. Therefore, in this research, model and simple application are made to evaluate service quality of PT. Pos Indonesia courier. <br /> <br /> This research explored the possibility of building two text mining model based on sentiment analysis and quality qualification. Machine learning approach with Support Vector Machine (SVM) algorithm is used as the basis of the model. There are 450 data used for sentiment analysis, with 150 of them are neutral, 150 negative data, and 150 positive. For quality classification, 450 data used with classification such as 150 for security, 150 for service fee, and 150 regarding delivery time. The accuracy of sentiment analysis model obtained at 96% while the quality classification model showed 97% of accuracy level. <br /> <br /> This research is also suggesting a simple prototype that based on both model and directly usable for real life cases currently experienced by PT. Pos Indonesia business division. The main process of the program is gathering tweets from Twitter feeds and visualized its output.<p align="justify">