Abstrak Bukunovel merupakan suatu karya sastra berbentuk prosa naratif yang panjang, yang memiliki rangkaian cerita tentang kehidupan seorang tokoh dan orang-orang di sekitarnya dengan menonjolkan sifat dan watak dari setiap tokoh pada novel tersebut. Dengan banyaknya review yang muncul dari pendapat pembaca, maka semakin sulit untuk menemukan review yang sesuai dengan pilihan konsumen. Ini yang menjadi suatu permasalahan yang dimana, konsumen tidak selalu menerima review dari pembaca. Untuk memecahkan permasalahan tersebut, diperlukan sebuah metode yang dapat memudahkan untuk menganalisis terkait dengan review tersebut. Maka dari itu, solusi yang akan dilakukan yaitu dengan menerapkan klasifikasi sentiment analysis. Sentiment analysis merupakan penambangan kontekstual data berupa teks, yang bertujuan untuk menganalisa berbagai pendapat atau opini berupa isu, komentar, dan lain-lain terhadap suatu objek atau permasalahan oleh seseorang yang dimana nilai tersebut akan memiliki nilai positif atau negatif. Penelitian ini, memiliki beberapa tujuan yaitu, untuk mengetahui performansi pada sistem klasifikasi Support Vector Machine (SVM) yang dibangun. Kedua, untuk mengetahui performansi pada pembobotan fitur Term Frequency-Inverse Document Frequency (TF-IDF) dan seleksi fitur Chi Square. Ketiga, Untuk mengetahui performansi pada pembobotan fitur Term Frequency (TF) dan seleksi fitur Chi Square. Dari hasil eksperimen, diperoleh bahwa hasil perfomansi terbaik untuk klasifikasi sentiment analysis pada review buku novel berbahasa Inggris, yaitu pada penggunaan kernel Gaussian RBF untuk setiap kedua pembobotan fitur dengan seleksi fitur yang digunakan dengan nilai performansi sebesar 74.2%. Kata kunci: Support Vector Machine (SVM), Sentiment Analysis, review, Term Frequency-Inverse Document Frequency (TF-IDF), Term Frequency (TF), Chi Square Abstract The novel book is a literary work in the form of a long narrative prose, which has a series of stories about the life of a character and people around him by highlighting the nature and character of each character in the novel. With so many reviews that arise from the opinions of readers, it is increasingly difficult to find reviews that are in accordance with consumer choice. This is a problem in which, consumers do not always receive reviews from readers. To solve this problem, we need a method that can make it easy to analyze related to the review. Therefore, the solution that will be carried out is by applying a sentiment analysis classification. Sentiment analysis is a contextual mining of data in the form of text, which aims to analyze various opinions or opinions in the form of issues, comments, etc. on an object or problem by someone whose value will have a positive or negative value. This study, has several objectives, namely, to determine the performance of the Support Vector Machine (SVM) classification system that was built. Second, to find out the performance in weighting the Term Frequency-Inverse Document Frequency (TF-IDF) feature and Chi Square feature selection. Third, to determine the performance of the Term Frequency (TF) weighting and Chi Square feature selection. From the experimental results, it was found that the best performance results for the classification of sentiment analysis in the review of English-language novel books, namely the use of Gaussian RBF kernels for each of the two weighting features with feature selection used with a performance value of 74.2%. Keywords: Support Vector Machine (SVM), Sentiment Analysis, review, Term Frequency-Inverse Document Frequency (TF-IDF), Term Frequency (TF), Chi Square
[1]
S. Vijayarani,et al.
Preprocessing Techniques for Text Mining-An Overview Dr
,
2015
.
[2]
Vimalkumar B. Vaghela,et al.
Sentiment Analysis using Support Vector Machine based on Feature Selection and Semantic Analysis
,
2016
.
[3]
Pablo Gamallo,et al.
Comparing Supervised Machine Learning Strategies and Linguistic Features to Search for Very Negative Opinions
,
2018,
Inf..
[4]
John Blitzer,et al.
Biographies, Bollywood, Boom-boxes and Blenders: Domain Adaptation for Sentiment Classification
,
2007,
ACL.
[5]
Muhammad Azeem,et al.
3D Shape Retrieval using Bag of Word Approaches
,
2019,
2019 2nd International Conference on Computing, Mathematics and Engineering Technologies (iCoMET).
[6]
Ahmad Fathan Hidayatullah,et al.
ANALISIS SENTIMEN DAN KLASIFIKASI KATEGORI TERHADAP TOKOH PUBLIK PADA TWITTER
,
2014
.
[7]
M. Ali Fauzi,et al.
Klasifikasi Teks Bahasa Indonesia Pada Dokumen Pengaduan Sambat Online Menggunakan Metode K-Nearest Neighbors Dan Chi-square
,
2017
.
[8]
Laura Ferrari,et al.
A Comparison between Preprocessing Techniques for Sentiment Analysis in Twitter
,
2016,
KDWeb.
[9]
Esha Tyagi,et al.
Sentiment Analysis of Product Reviews using Support Vector Machine Learning Algorithm
,
2017
.
[10]
Sentiment Analysis Berbahasa Indonesia Menggunakan Improved Multinomial Naive Bayes
,
2015
.
[11]
Kusrini,et al.
Komparasi Teknik Klasifikasi Teks Mining Pada Analisis Sentimen
,
2016
.
[12]
Sarit Chakraborty,et al.
An Improved Text Sentiment Classification Model Using TF-IDF and Next Word Negation
,
2018,
ArXiv.
[13]
Much Aziz Muslim,et al.
Improve the Accuracy of Support Vector Machine Using Chi Square Statistic and Term Frequency Inverse Document Frequency on Movie Review Sentiment Analysis
,
2019,
Scientific Journal of Informatics.
[14]
Adiwijaya Adiwijaya,et al.
Klasifikasi Sentiment Analysis pada Review Film Berbahasa Inggris dengan Menggunakan Metode Doc2Vec dan Support Vector Machine (SVM)
,
2018
.
[15]
Fiqih Munakahat,et al.
DAFTAR PUSTAKA
,
2002
.