Method comparison of Naïve Bayes, Logistic Regression, and SVM for Analyzing Movie Reviews
-
Published:2023-03-29
Issue:4
Volume:4
Page:
-
ISSN:2685-3310
-
Container-title:Building of Informatics, Technology and Science (BITS)
-
language:
-
Short-container-title:bits
Author:
Aziz Muhammad Maulidan,Purbalaksono Mahendra Dwifebri,Adiwijaya Adiwijaya
Abstract
A film can be categorized as a successful film based on the reviews given by the critics. The reviews can range from professional critics to public reviews from the general audience. Due to a large number of reviews and opinions on a film, this study aims to create a sentiment analysis model and compare the methods used to analyze datasets from a movie review. Sentiment Analysis is a method for studying and analyzing opinions, then classifying these opinions into several classes. This research will use the Naïve Bayes method, Logistic Regression, and Support Vector Machine (SVM) to analyze film review data. The film review dataset used is a collection of film reviews taken from the Rotten Tomatoes website and will be pre-processed before implementing the Naïve Bayes, Logistic Regression, and SVM methods. The SVM classifier with 80:20 data splitting has the best performance, with a result of 99.4% accuracy score and 93.5% F1 score.
Publisher
Forum Kerjasama Pendidikan Tinggi (FKPT)
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献