Sentiment Analysis on Indonesian Movie Review Using KNN Method With the Implementation of Chi-Square Feature Selection

Imam Prayoga (Telkom University, Bandung)
Mahendra Dwifebri Purbolaksono (Telkom University, Bandung)
Adiwijaya Adiwijaya (Telkom University, Bandung)

JURNAL MEDIA INFORMATIKA BUDIDARMA
Vol 7, No 1 (2023): Januari 2023

The advancement and development of the internet is used by the people to support various sectors, one of which is the film industry. Nowadays, people can easily access various movies from available sites. This convenience had led to many reviews about a movie that can be obtained easily. This movie review is very influential on the variety of movies. Freedom of expression on the internet, makes the reviews of a movie vary. For this reason, it is necessary to analyze the sentiment of he movie reviews that are positive or negative. In this research, a sentiment analysis model is build using chi-square selection feature with the KNN algorithm. The final result of this research is able to provide the best classification model with the implementation of stemming. The value of k = 267 in selectkbest at the feature selection stage using chi-square, and using the value of K = 11 in the KNN parameter. This model produces f1 score value of 86.98%.