TOPIC MODELLING DAN TOPIC GENERATION PADA REVIEW GAME STEAM DENGAN FINE-TUNING BERTOPIC DAN QWEN2.5

DANI, MUHAMMAD RIZALUL FIQRI SYAH and Satria, Hadipurnawan (2025) TOPIC MODELLING DAN TOPIC GENERATION PADA REVIEW GAME STEAM DENGAN FINE-TUNING BERTOPIC DAN QWEN2.5. Undergraduate thesis, Sriwijaya University.

Preview	Image RAMA_55201_09021282126114_0018048003_cover.jpg - Cover Image Available under License Creative Commons Public Domain Dedication. Download (138kB) | Preview
	Text RAMA_55201_09021282126114.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_55201_09021282126114_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (20MB) | Request a copy
	Text RAMA_55201_09021282126114_0018048003_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (852kB)
	Text RAMA_55201_09021282126114_0018048003_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (677kB) | Request a copy
	Text RAMA_55201_09021282126114_0018048003_03.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (325kB)
	Text RAMA_55201_09021282126114_0018048003_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (750kB) | Request a copy
	Text RAMA_55201_09021282126114_0018048003_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (156kB) | Request a copy
	Text RAMA_55201_09021282126114_0018048003_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (79kB) | Request a copy
	Text RAMA_55201_09021282126114_0018048003_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (5kB) | Request a copy

Abstract

This study aims to develop and evaluate a Topic Modelling and Topic Generation system using user review data from the Steam platform. The research utilizes the BERTopic model to cluster reviews into topics in an unsupervised manner, and the Qwen2.5 Large Language Model to generate more informative topic representations. Various algorithm combinations covering embedding, dimensionality reduction, and clustering were tested to determine the optimal configuration using evaluation metrics such as C_V, U_Mass, and C_NPMI. Additionally, the quality of topics generated by Qwen2.5 was evaluated through Cosine Similarity scores. The results indicate that certain algorithm combinations produce more accurate and representative topic clusters. The developed system also features a user-friendly interface for result exploration. This research contributes to text review processing in the gaming domain and serves as a foundation for future studies in opinion analysis and natural language processing.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Topic Modelling, Large Language Model, BERTopic, Qwen2.5, Pemrosesan Bahasa Alami, Generasi Topik
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:	Muhammad Rizalul Fiqri Syah Dani
Date Deposited:	04 Jun 2025 03:21
Last Modified:	04 Jun 2025 03:21
URI:	http://repository.unsri.ac.id/id/eprint/175059

Actions (login required)

View Item