Jurnal Mandiri IT
Vol. 14 No. 1 (2025): July: Computer Science and Field.

K-Means clustering analysis of public satisfaction with 50% electricity tariff reduction

Harahap, Muhammad Fitrah Affandi (Unknown)
Hasugian, Abdul Halim (Unknown)



Article Info

Publish Date
15 Jul 2025

Abstract

At the beginning of 2025, the Indonesian government implemented a policy to reduce electricity tariffs by 50% for household customers with power capacities of up to 2,200 VA. This policy aims to boost public purchasing power and stimulate economic growth, particularly among lower-middle-income groups. However, public responses to the policy have been varied and widely expressed on social media, especially on platform X (formerly known as Twitter). This study aims to evaluate public satisfaction with the electricity tariff reduction policy through sentiment analysis on social media X using the K-Means Clustering method. Data were collected through a crawling process using specific relevant keywords, followed by preprocessing steps such as cleansing, case folding, tokenizing, stemming, and conversion into numerical form using TF-IDF. The clustering results show that Cluster 1 dominates with 662 tweets (68.74%), predominantly reflecting positive sentiment, indicating that the majority of the public responded favorably to the 50% electricity tariff reduction policy. Cluster 2 consists of 165 tweets (17.13%) expressing negative sentiment, suggesting that some members of the public voiced concerns or dissatisfaction with the policy. Meanwhile, Cluster 0 includes 136 tweets (14.12%) containing neutral sentiment, representing moderate responses without a strong stance. These findings indicate that, overall, the policy received a generally positive reception from the public, although there are still critical and neutral perspectives.

Copyrights © 2025






Journal Info

Abbrev

Mandiri

Publisher

Subject

Computer Science & IT Library & Information Science Mathematics

Description

The Jurnal Mandiri IT is intended as a publication media to publish articles reporting the results of Computer Science and related ...