FEATURE AND SERVICE QUALITY IMPROVEMENT OF JENIUS DIGITAL BANK USING TEXT MINING METHOD

• Main Author: Ramadhan Sugiantoro, Nivan
• Format: Theses
• Language: Indonesia
• Subjects: Manajemen umum
• Online Access: https://digilib.itb.ac.id/gdl/view/68667
• Institution: Institut Teknologi Bandung

The high development of the internet and technology also impacts banking with the existence of digital banking. Jenius has problems with poor online reviews. To maintain its existence in the digital industry and maintain customer loyalty, Bank Jenius tries to take advantage of customer reviews to improve the features and quality of its services. This study tries to reveal public sentiment, public opinion, and service quality as reflected in the Google Play review of the Jenius application. Public sentiment analysis and service quality analysis are carried out using a sentiment analysis classification algorithm. Public opinion is done by using topic modeling using the LDA algorithm. The best model generated by sentiment analysis is 87.13% using a deep learning algorithm. Meanwhile, the best result for service quality model is the k-NN algorithm with a value of 48%. From the model obtained, an analysis of the results was carried out. There are 7525 (71.2%) positive sentiments and 2689 (25.54%) negative sentiments in reviews for Jenius. Sentiment analysis is divided into analyzing the comments that get the most thumbs up and analyzing each star rating. The positive sentiment that is most often discussed is the ease of using the application. Negative sentiments talk about slow apps, problems logging in to apps, credit issues, apps that stop suddenly, verification issues, and security issues. In the analysis of service quality, four dimensions have positive sentiments, and three have negative sentiments. Positive sentiment is found in tangible, responsiveness, and efficiency dimensions. Meanwhile, negative sentiment is found in the dimensions of reliability, assurance, and empathy. The topic analysis resulted in 15 topics. Topics with negative sentiments are login, video call verification, device issues, promotions, app performance, transfer fees, and data privacy. In contrast, topics with positive sentiments are topics about transactions, customer satisfaction, savings, payments, consumer experience, Jenius usefulness, and payment history. From various analyzes, several problems were selected, namely problems getting into the application, slow applications, verification difficulties, and device problems. The solutions to these problems are UI/UX improvements and server upgrades to overcome slow application and application login problems, implementing AI-based verification for data verification, and adding device facilities for device problems.