ANALISIS SENTIMEN BERDASARKAN ULASAN UNTUK REKOMENDASI COFFEE SHOP DI DAERAH ISTIMEWA YOGYAKARTA MENGGUNAKKAN METODE TF-IDF DAN SUPPORT VECTOR MACHINE (SVM)

Larasati, Ahlida Sabila (2024) ANALISIS SENTIMEN BERDASARKAN ULASAN UNTUK REKOMENDASI COFFEE SHOP DI DAERAH ISTIMEWA YOGYAKARTA MENGGUNAKKAN METODE TF-IDF DAN SUPPORT VECTOR MACHINE (SVM). Other thesis, UPN Veteran Yogyakarta.

	Text Abstrak_Ahlida Sabila.pdf Download (281kB)
	Text Cover Skripsi_Ahlida Sabila.pdf Download (254kB)
	Text Daftar Isi_Ahlida Sabila.pdf Download (243kB)
	Text Daftar Pustaka_Ahlida Sabila.pdf Download (251kB)
	Text Halaman Pengesahan Pembimbing_Ahlida.pdf Download (311kB)
	Text Halaman Pengesahan Penguji_Ahlida.pdf Download (316kB)
	Text SKRIPSI FINAL AHLIDA SABILA LARASATI.pdf Restricted to Repository staff only Download (7MB)

Abstract

vi
ABSTRAK
Daerah Istimewa Yogyakarta sebagai salah satu provinsi besar di Indonesia memiliki
banyak coffee shop yang menawarkan beragam suasana, jenis kopi, dan layanan unik. Hal ini
dapat dilihat dari banyaknya pilihan coffee shop di Daerah Istimewa Yogyakarta yang
menyajikan kopi dengan cita rasa khas. Ulasan pelanggan memiliki peran penting dalam
membentuk citra dan reputasi sebuah coffee shop, dan banyak orang mengandalkan review dari
pengguna Google Maps untuk membantu mereka memilih tempat yang tepat. Oleh karena itu,
diperlukan analisis sentimen yang mampu mengklasifikasikan ulasan tersebut ke dalam
kategori sentimen positif, negatif, atau netral. Hasil dari analisis ini dapat digunakan untuk
memberikan rekomendasi coffee shop di Daerah Istimewa Yogyakarta. Support Vector Machine
(SVM) merupakan algoritma yang sangat bergantung pada kualitas fitur yang diberikan. Jika
fitur yang digunakan tidak relevan, model Support Vector Machine (SVM) tidak akan dapat
bekerja dengan baik, namun Term Frequency-Inverse Document Frequency (TF-IDF) bisa
mengekstrak fitur yang lebih bermakna dari teks, sehingga SVM dapat melakukan klasifikasi
dengan lebih akurat.
Penelitian ini menggunakkan metode Support Vector Machine (SVM) yang
dikombinasikan dengan metode Term Frequency-Inverse Document Frequency (TF-IDF).
Analisis sentiment pada penelitian ini terdiri dari tiga kelas yaitu kelas positif, kelas negatif,
dan kelas netral. Tahapan penelitian ini dimulai dengan pengumpulan data ulasan dari google
maps dengan cara scraping, kemudian dilakukan tahapan preprocessing. Data hasil dari
preprocessing selanjutnya diberi label menjadi tiga kelas, yaitu positif, negatif dan netral,
selanjutnya tahap ekstrasi fitur dengan TF-IDF, kemudian dilakukan tahap pemodelan dengan
metode SVM. Model yang telah dibangun dilakukan pengujian dengan tabel confusion matrix.
Dari tabel confusion matrix tersebut akan diperoleh nilai akurasi, presisi, serta recall.
Berdasarkan hasil pengujian dengan menggunakkan 1484 data dengan perbandingan
data latih dan data uji yaitu 80% : 20%, dimana untuk data latih sebanyak 1.187 data dan data
uji sebanyak 297 data, dengan pengujian menggunakan tabel confusion matrix dihasilkan nilai
akurasi sebesar 85.86%, nilai presisi sebesar 86.24%, serta nilai recall sebesar 97.89%.
Kata kunci : Coffee Shop, Analisis Sentimen, TF-IDF, SVM, Confusion Matrix.
vii
ABSTRACT
The Special Region of Yogyakarta as one of the largest provinces in Indonesia has many
coffee shops that offer a variety of atmospheres, types of coffee, and unique services. This can
be seen from the many choices of coffee shops in the Special Region of Yogyakarta that serve
coffee with distinctive flavors. Customer reviews play an important role in shaping the image
and reputation of a coffee shop, and many people rely on reviews from Google Maps users to
help them choose the right place. Therefore, a sentiment analysis is needed that is able to
classify these reviews into positive, negative, or neutral sentiment categories. The results of
this analysis can be used to provide recommendations for coffee shops in the Special Region of
Yogyakarta. Support Vector Machine (SVM) is an algorithm that is highly dependent on the
quality of the features provided. If the features used are not relevant, the Support Vector
Machine (SVM) model will not be able to work well, but Term Frequency-Inverse Document
Frequency (TF-IDF) can extract more meaningful features from the text, so that SVM can
classify more accurately.
This study uses the Support Vector Machine (SVM) method combined with the Term
Frequency-Inverse Document Frequency (TF-IDF) method. Sentiment analysis in this study
consists of three classes, namely positive class, negative class, and neutral class. The stages of
this study began with collecting review data from Google Maps by scraping, then the
preprocessing stage was carried out. The data from the preprocessing results were then labeled
into three classes, namely positive, negative and neutral, then the feature extraction stage with
TF-IDF, then the modeling stage was carried out with the SVM method. The model that has
been built was tested with a confusion matrix table. From the confusion matrix table, the
accuracy, precision, and recall values will be obtained.
Based on the test results using 1484 data with a comparison of training data and test
data of 80%: 20%, where for training data as many as 1,187 data and test data as many as 297
data, testing using the confusion matrix table produced an accuracy value of 85.86%, a
precision value of 86.24%, and a recall value of 97.89%.
Keywords: Coffee Shop, Sentiment Analysis, TF-IDF, SVM, Confusion Matrix.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Coffee Shop, Sentiment Analysis, TF-IDF, SVM, Confusion Matrix.
Subjects:	Z Bibliography. Library Science. Information Resources > ZA Information resources
Divisions:	Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	Eko Yuli
Date Deposited:	24 Oct 2024 06:42
Last Modified:	24 Oct 2024 06:42
URI:	http://eprints.upnyk.ac.id/id/eprint/41470

Actions (login required)

View Item