ASPECT-BASED SENTIMENT ANALYSIS (ABSA) WONDR BY BNI MENGGUNAKAN CNN, LSTM, SVM, DAN NAIVE BAYES

RARAS, FATIMAH WANUDYA and Meiriza, Allsela (2025) ASPECT-BASED SENTIMENT ANALYSIS (ABSA) WONDR BY BNI MENGGUNAKAN CNN, LSTM, SVM, DAN NAIVE BAYES. Undergraduate thesis, Sriwijaya University.

Preview	Image RAMA_57201_09031282126068_cover.jpg - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (163kB) | Preview
	Text RAMA_57201_09031282126068.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (8MB) | Request a copy
	Text RAMA_57201_09031282126068_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (13MB) | Request a copy
	Text RAMA_57201_09031282126068_0013058302_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (927kB)
	Text RAMA_57201_09031282126068_0013058302_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (307kB) | Request a copy
	Text RAMA_57201_09031282126068_0013058302_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_57201_09031282126068_0013058302_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (5MB) | Request a copy
	Text RAMA_57201_09031282126068_0013058302_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (11kB) | Request a copy
	Text RAMA_57201_09031282126068_0013058302_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (161kB) | Request a copy
	Text RAMA_57201_09031282126068_0013058302_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy

Abstract

Wondr by BNI is a banking innovation since its launch in July 2024, has garnered widespread user attention with millions of downloads and thousands of reviews. These user reviews can be thoroughly analyzed using an Aspect-Based Sentiment Analysis (ABSA) approach. This study aims to apply ABSA to reviews of the Wondr app, focusing on four key aspects of usability according to Nielsen: learnability, efficiency, error, and satisfaction. A total of 5,500 review data points were analyzed using CNN, LSTM, SVM, and Naive Bayes algorithms, with feature extraction via Word2Vec. The study results show that the error aspect is the most frequently discussed by users, with the majority of sentiments being negative. Conversely, the satisfaction aspect is dominated by positive sentiments. From the model performance evaluation, the CNN model demonstrated the best overall performance, with the highest accuracy and F1-score across most aspects, and effectively leveraged Word2Vec representations to understand the context of user reviews. The LSTM model showed slightly lower performance than CNN. SVM delivered good results, while the Naive Bayes model consistently showed the lowest performance.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Aspect-Based Sentiment Analysis (ABSA), Wondr by BNI, CNN, LSTM, SVM, Naive Bayes.
Subjects:	Q Science > Q Science (General) > Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation.
Divisions:	09-Faculty of Computer Science > 57201-Information Systems (S1)
Depositing User:	Fatimah Wanudya Raras
Date Deposited:	21 May 2025 07:35
Last Modified:	21 May 2025 07:35
URI:	http://repository.unsri.ac.id/id/eprint/173512

Actions (login required)

View Item