International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 176 - Number 26 |
Year of Publication: 2020 |
Authors: Penny Dwi Harnaning, Muhammad Subali, Karmilasari |
10.5120/ijca2020920236 |
Penny Dwi Harnaning, Muhammad Subali, Karmilasari . Application of Naïve Bayes Algorithm for Measuring the Suitability of the Work Position in Ministry of Home Affairs. International Journal of Computer Applications. 176, 26 ( May 2020), 28-34. DOI=10.5120/ijca2020920236
The Ministry of Home Affairs a ministry which is under and responsible to the President through the Minister, who has the task of holding affairs in the field of internal government to assist the President in organizing the governmence of the country. In carrying out governmental tasks of domestic government, the Ministry of Home Affairs requires qualified State Civil Apparatuses and their placement must be in accordance with their competencies. Until now there are still problems with the placement of employees in positions that are not in line with their competencies. Data mining in this study was implemented to measure the level of suitability of the employees to the positions they occupy by involving large amounts of data, the technique used for classification is the Naïve Bayes algorithm which is used to determine the extent of suitability between employees and the occupied positions. The attributes used consist of three attributes, namely the level of education, education and training, and rank / classfication. The object of this study are 4202 employee profile data consisting of 305 employees who hold the position of Administrator (III. A), 16 employees who occupy the position of Administrator (III. B), 806 employees who occupy Supervisory positions (IV. A), 82 employees occupying the Supervisory position (IV. B), and 2993 employees who hold the Implementing position. Accuracy of suitability between employees and the position they occupy is based on testing of the results of the classification of the Naïve Bayes algorithm using 90% training data and 10% testing data is 83%.