期刊名称：International Journal on Electrical Engineering and Informatics
印刷版ISSN：2085-6830
出版年度：2019
卷号：11
期号：3
页码：474-484
DOI：10.15676/ijeei.2019.11.3.2
出版社：School of Electrical Engineering and Informatics
摘要：Interviewer falsification is an important issue faced by institutions conductingcensuses and surveys around the world, including Statistics Indonesia. This study discussesseveral methods to systematically detect interviewer falsification and validation using datamining techniques so that human supervisors can take further actions. After analyzing relevantfeatures and conducting experiments, the results showed that unsupervised classificationalgorithm using simple 2-means clustering achieved 70.5% accuracy, while the supervisedclassification using logistic regression improved the accuracy to 88.5%. A greater level ofaccuracy is still needed to be pursued in further research, but the current results are certainlybetter than the traditional method which has almost no falsification detection method at all.
关键词：Data Mining; Interviewer Falsification; Mobile Survey; Statistics
