Analisis sentimen data ulasan aplikasi KAI Access pada Google Play Store menggunakan metode naive bayes classifier

Ningsih, Erik Dwi (2022) Analisis sentimen data ulasan aplikasi KAI Access pada Google Play Store menggunakan metode naive bayes classifier. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 10611810000089-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (2MB)

Abstract

Aplikasi KAI Access merupakan aplikasi yang memberikan kemudahan bagi calon penumpang kereta api dalam melakukan pemesanan tiket secara daring tanpa harus mengantri di loket stasiun. Hal tersebut memberikan dampak positif terhadap masyarakat yang sering menggunakan jasa kereta api. Namun, dalam penggunaannya, aplikasi yang dapat diunduh melalui situs Google Play Store ini sering mendapatkan ulasan berupa kritik negatif dari pengguna. Berdasarkan penilaian pada situs Google Play Store, KAI Access mempunyai rating sebesar 4,2 dari skala 5, di mana nilai tersebut tergolong rendah dibandingkan dengan aplikasi e-ticket lainnya, seperti Traveloka sebesar 4,8 dan tiket.com sebesar 4,7. Hal tersebut melatarbelakangi dilakukannya peninjauan terkait pelayanan sistem aplikasi melalui ulasan yang diberikan oleh pengguna. Banyaknya data ulasan diperlukan adanya penerapan data mining untuk membantu menganalisis teks ulasan. Analisis yang dilakukan pada penelitian ini adalah analisis sentimen menggunakan metode Naïve Bayes Classifier dan pembobotan Raw-TF. Hasil penelitian menunjukkan bahwa persentase ulasan dengan sentimen positif sebesar 58%, di mana nilai tersebut lebih tinggi dibandingkan dengan ulasan sentimen negatif sebesar 42%. Pada hasil perhitungan ketepatan klasifikasi diperoleh accuracy sebesar 67,86%, sensitivity sebesar 95,39%, dan specificity sebesar 41,52%. Sedangkan ketepatan klasifikasi pada data testing diperoleh accuracy sebesar 69,51%, sensitivity sebesar 95,89%, dan specificity sebesar 40,48%. Penelitian ini diharapkan dapat memberikan informasi kepada pihak KAI Access terkait peningkatan layanan dalam hal pembayaran dan meminimalisasi terjadinya error dalam melakukan pemesanan tiket melalui aplikasi.
================================================================================================================================
The KAI Access application is an application that makes it easier for prospective train passengers to purchase tickets online without having to queue at the station counter. This has a positive impact on people who frequently use train services. However, in its use, the application, which can be downloaded through the Google Play Store, often receives reviews in the form of negative criticism from users. Based on the rating on the Google Play Store, KAI Access has a rating of 4.2 out of 5, which is relatively low compared to other e-ticket applications, such as Traveloka with a rating of 4.8 and ticket.com with a rating of 4.7. This has prompted a review of the application system services based on reviews provided by users. The large amount of review data requires the application of data mining to help analyze text reviews. The analysis conducted in this study is sentiment analysis using the Naïve Bayes Classifier method and Raw-TF weighting. The results showed that the percentage of reviews with positive sentiment was 58%, which was higher than the percentage of reviews with negative sentiment, at 42%. The classification accuracy results obtained an accuracy of 75.70%, sensitivity of 93.90%, and specificity of 50.67% for the training data, while the classification accuracy for the testing data obtained an accuracy of 75.08%, sensitivity of 94.49%, and specificity of 47.79%. This research is expected to provide information to KAI Access regarding service improvements, particularly in terms of payment, and to minimize the occurrence of errors when ordering tickets through the application.

Item Type:	Thesis (Other)
Additional Information:	519.53 Nin a-1
Uncontrolled Keywords:	Analisis Sentimen, KAI Access, Naïve Bayes Classifier, Pembobotan Raw-T, Sentiment Analysis, KAI Access, Naïve Bayes Classifier, Raw-TF Weighting
Subjects:	Q Science > QA Mathematics > QA278.55 Cluster analysis
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	magang .
Date Deposited:	06 Oct 2026 06:33
Last Modified:	06 Oct 2026 06:33
URI:	http://repository.its.ac.id/id/eprint/145326

Actions (login required)

View Item