This Author published in this journals
All Journal Jurnal Mantik
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Demographic Attribute Selection Model For Prediction Of Election Participation Using Decision Tree Linda Kushernawati; Arif Senja Fitrani; Metatia Intan Mauliana
Jurnal Mantik Vol. 6 No. 2 (2022): August: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mantik.v6i2.2848

Abstract

Implementing a democratic general election is expected to produce people's representatives who can channel the people's aspirations. Demographic data is information that discusses a group of people with several related attributes and involves many factors. In this study, we will relate the relationship between the implementation of elections and the condition of demographic data with a benchmark for the form of public participation in the election. By utilizing 2019 election data and Bangkalan Regency demographic data from the Central Statistics Agency (BPS), it is expected to determine the relationship between the two conditions of the dataset on the form of public participation at the polling station (TPS) level. By starting with the Preprocessing step, it implement a classification method with the Decision Tree (DT) algorithm to predict community presence at the polling station level. There are three versions of the dataset that will be used in modeling, namely initial data that has not been selected for attributes (version 1), data that has been chosen using correlation-based attribute selection (version 2), and data that has been selected using chi-square attributes ( version 3). The results show version 1 with a prediction of 81%, followed by version 2 with a prediction of 81%, and the last is version 3 with a prediction of 70%. The detachment model's formation with the selection attribute has a different impact, and the relationship between the election dataset and demographics has a significant effect, as indicated by the prediction results of version 2.