Success factors for citizen-based government decision making using K-means fuzzy learning vector quantization

Abstract

The Indonesian government often needs assistance in making citizen-based decisions, for example selecting work program plans. Residents have their criteria in the forum to choose a work program plan. This study proposes the K-means fuzzy learning vector quantization (FLVQ) methods to select citizen-based government decision-making criteria. The K-means FLVQ method has never been used to assist government decision-making. However, citizen criteria can be a success factor for government decision-making. The selection of criteria begins with data collection from forum participants. The results of data collection get 11 criteria. Then, the K-means FLVQ method carries out labeling and classification. The addition of the K-means process in the selection criteria can provide optimal results. Citizens can give assessment criteria freely. Then the assessment of citizens is classified by FLVQ. The classification results obtained seven criteria, namely: i) urgency, ii) sustainability, iii) priority, vi) usability, v) prosperity, vi) comfortability, and vii) artistic. Governments can use these criteria to make decisions about planned work programs. The criteria selection algorithm was also evaluated using the confusion matrix method with an accuracy of 88% and an error of 12%.