ANALISIS SENTIMEN ULASAN APLIKASI SPOTIFY DI GOOGLE PLAY STORE MENGGUNAKAN MODEL BIDIRECTIONAL ENCODER REPRESENTATIONS FROM TRANSFORMER (BERT)

ZIKRI, ANHARUL and Yusliani, Novi and Marieska, Mastura Diana (2025) ANALISIS SENTIMEN ULASAN APLIKASI SPOTIFY DI GOOGLE PLAY STORE MENGGUNAKAN MODEL BIDIRECTIONAL ENCODER REPRESENTATIONS FROM TRANSFORMER (BERT). Undergraduate thesis, Sriwijaya University.

	Image RAMA_55201_09021282126034_cover.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (962kB)
	Text RAMA_55201_09021282126034.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) | Request a copy
	Text RAMA_55201_09021282126034_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (8MB) | Request a copy
	Text RAMA_55201_09021282126034_0008118205_0021038607_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (5MB)
	Text RAMA_55201_09021282126034_0008118205_0021038607_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (5MB) | Request a copy
	Text RAMA_55201_09021282126034_0008118205_0021038607_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_55201_09021282126034_0008118205_0021038607_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_55201_09021282126034_0008118205_0021038607_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (4MB) | Request a copy
	Text RAMA_55201_09021282126034_0008118205_0021038607_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_55201_09021282126034_0008118205_0021038607_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (44kB) | Request a copy

Abstract

Spotify, with 574 million active users and 226 million premium subscribers across 184 countries as of Q3 2023, represents a significant force in music streaming. While Google Play Store ratings provide initial app quality indicators, they fail to capture comprehensive user experiences. This research employs Bidirectional Encoder Representations from Transformers (BERT) to conduct sentiment analysis on 52,702 Spotify app reviews from Google Play Store, categorizing them as positive, negative. Eight scenario were tested, varying learning rates (2e-5 and 2e-6) and epoch counts (5, 10, 15, and 20). Results demonstrate optimal performance with 92.92% accuracy using a 2e-6 learning rate and 10 epochs, confirming that appropriate hyperparameter selection creates stable models while preventing overfitting issues.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Analisis Sentimen, BERT, Spotify, Ulasan Pengguna, Google Play Store.
Subjects:	Q Science > Q Science (General) > Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation.
Divisions:	09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User:	Anharul Zikri
Date Deposited:	14 May 2025 07:31
Last Modified:	14 May 2025 07:31
URI:	http://repository.unsri.ac.id/id/eprint/172359

Actions (login required)

View Item