Analisis Sentimen Berbasis Aspek Pada Ulasan Aplikasi Midi Kriing Menggunakan Support Vector Machine (SVM)

Fahmi, Rohmat Ubaidillah (2025) Analisis Sentimen Berbasis Aspek Pada Ulasan Aplikasi Midi Kriing Menggunakan Support Vector Machine (SVM). Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21082010099-cover.pdf Download (1MB)
	Text (BAB I) 21082010099-bab1.pdf Download (31kB)
	Text (BAB II) 21082010099-bab2.pdf Restricted to Repository staff only until 23 May 2027. Download (202kB) \| Request a copy
	Text (BAB III) 21082010099-bab3.pdf Restricted to Repository staff only until 23 May 2027. Download (666kB) \| Request a copy
	Text (BAB IV) 21082010099-bab4.pdf Restricted to Repository staff only until 23 May 2027. Download (5MB) \| Request a copy
	Text (BAB V) 21082010099-bab5.pdf Download (11kB)
	Text (Daftar Pustaka) 21082010099-daftarpustaka.pdf Download (97kB)
	Text (Lampiran) 21082010099-lampiran.pdf Restricted to Repository staff only until 23 May 2027. Download (210kB) \| Request a copy

Abstract

Midi Kriing is an online shopping application developed by PT Midi Utama Indonesia Tbk (Alfamidi) to meet customer needs. To improve service quality and user experience, sentiment analysis of the Midi Kriing application has become crucial. Sentiment analysis was performed on three main aspects of user reviews of the Midi Kriing application: products, services, and application functionality, using the Support Vector Machine (SVM) algorithm. The selection of these three aspects was a result of an agreement between the researchers and partners, focusing on the company's efforts to optimize the quality and features of the application that directly affect the user experience. This research produced three separate sentiment classification models, each for the product, service, and application functionality aspects. In the experiment, 36 scenarios were conducted to optimize the results of each aspect model. The best results from the experiment are as follows: 1) Product aspect model with an F1 Score of 0.75 and Accuracy of 0.86, 2) Service aspect model with an F1 Score of 0.84 and Accuracy of 0.86, and 3) Application functionality aspect model with an F1 Score of 0.84 and Accuracy of 0.89. The best model from the three aspects was implemented on a website using the Flask framework. This website can predict sentiment based on input sentence reviews, classify reviews in .csv format, and display classification results along with aspect distribution graphs and ratings. The choice of the SVM algorithm was due to its advantages in handling large volumes of text data with complex features and its ability to separate classes with an optimal margin, especially in non-linear data. Compared to other algorithms such as Naïve Bayes and k-Nearest Neighbor, SVM has proven to be superior in dealing with data imbalance issues and producing models with better performance. Therefore, this research provides in-depth information regarding user sentiment toward the Midi Kriing application, which can be used by Alfamidi to better understand customer needs and improve the quality of service and user experience with the application.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Arifiyanti, Amalia Anjani	NIDN0712089201	amalia_anjani.fik@upnjatim.ac.id
Thesis advisor	Sugata, Tri Luhur Indayanti	NIP199206162024062001	tri.luhur.fasilkom@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)
T Technology > T Technology (General) > T58.6-58.62 Management Information Systems

Divisions:

Faculty of Computer Science > Departemen of Information Systems

Depositing User:

Rohmat Ubaidillah Fahmi

Date Deposited:

23 May 2025 06:08

Last Modified:

23 May 2025 06:08

URI:

https://repository.upnjatim.ac.id/id/eprint/36466