Pengaruh RFE Terhadap Logistic Regression Dan Support Vector Machine Pada Analisis Sentimen Hotel Shangri-La Surabaya

Herza, Fakhri Maulana (2024) Pengaruh RFE Terhadap Logistic Regression Dan Support Vector Machine Pada Analisis Sentimen Hotel Shangri-La Surabaya. Undergraduate thesis, UPN Veteran JAWA TIMUR.

	Text (Cover) 20081010039-cover.pdf Download (1MB)
	Text (Bab 1) 20081010039-bab1.pdf Download (108kB)
	Text (Bab 2) 20081010039-bab2.pdf Restricted to Repository staff only until 20 September 2026. Download (319kB)
	Text (Bab 3) 20081010039-bab3.pdf Restricted to Repository staff only until 20 September 2026. Download (584kB)
	Text (Bab 4) 20081010039-bab4.pdf Restricted to Repository staff only until 20 September 2026. Download (1MB)
	Text (Bab 5) 20081010039-bab5.pdf Download (12kB)
	Text (Daftar Pustaka) 20081010039-daftar pustaka.pdf Download (90kB)

Abstract

Sentiment analysis is one of the essential tools in the tourism industry for understanding guest responses and experiences regarding hotel services. Previous guest reviews play a crucial role in shaping the perceptions of potential guests about the quality of facilities and the attractiveness of the hotel they plan to visit. However, one of the main challenges in sentiment analysis is selecting the most relevant features to improve model prediction performance. Many reviews contain diverse information, and not all words or features significantly contribute to distinguishing between positive and negative sentiment. In this context, the Recursive Feature Elimination (RFE) method has the potential to optimize feature selection by eliminating less informative features, thus expected to improve the accuracy of Logistic Regression and Support Vector Machine (SVM) models in sentiment analysis. Therefore, this study focuses on the impact of RFE application on the performance of both models, particularly in analyzing guest reviews at the Shangri-La Hotel Surabaya. The data used in this study consists of 3719 reviews. The test results show that in the Logistic Regression model using RFE, there was a significant improvement in performance in terms of precision, sensitivity, F1 Score, and accuracy, with an average increase of 9%. Moreover, for the Support Vector Machine model using RFE, the performance improvement was even more significant, with an average increase of 14%. These findings indicate that the application of RFE can effectively enhance the predictive quality of both models in the context of hotel review sentiment analysis.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Rahmat, Basuki	NIDN0023076907	basukirahmat.if@upnjatim.ac.id
Thesis advisor	Al Haromainy, M. Muharrom	NIDN0701069503	muhammad.muharrom.if@upnjatim.ac.id

Subjects:

Q Science > Q Science (General)
T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Fakhri Maulana Herza

Date Deposited:

20 Sep 2024 02:56

Last Modified:

20 Sep 2024 02:56

URI:

https://repository.upnjatim.ac.id/id/eprint/29648