Journal of Applied Engineering and Technological Science Vol 1(2) 2020 : 103-112 103 DECISION SUPPORT SYSTEMS EMPLOYEE DISCIPLINE IDENTIFICATION USING THE SIMPLE MULTI ATTRIBUTE RATING TECHNIQUE (SMART) METHOD Rizer Fahlepi STIKes Hang Tuah Pekanbaru rizerf@htp.ac.id A B S T R A C T Discipline is a very important aspect to support the quality of these human resources. If there are insufficient or undisciplined resources, it will affect the quality of human resources. In its implementation the process of evaluating employee discipline is still done manually so it takes a long time. For this reason, a decision support system is needed to identify the level of discipline of staff and employees at STIKes and STMIK Hang Tuah Pekanbaru. The method used to develop this Decision Support Systems is Simple Multi Attribute Rating Technique (SMART) with 6 attributes, namely performance, warning letters, absenteeism, discipline, complience to regulations, and compliance to superior’s order. The final result are divided into 3 categories, namely Very Good, Enough, Do Coaching. From the application of this method, it was found that 120 people got Very Good evaluation results, 11 people got Enough results, and 2 people got the results of Do Coaching. Decision support system with SMART method can identify the level of discipline of staff and, to later be given guidance to staff who get evaluation results. Do Coaching to become more disciplined and improve the quality of human resources in STIKes and STMIK Hang Tuah Pekanbaru. Keywords: Decision Support Systems, SMART, Employee, Discipline.. 1. INTRODUCTION Work Discipline is the attitude and behavior of employees to comply with applicable regulations and adjust organizations to be based on self-awareness. Indicators of work discipline, namely the frequency of attendance at the office on weekdays and the accuracy of hours of entry and return, compliance with applicable regulations, compliance with specified work standards, and employee work ethics at the Institute (Thaief, Baharuddin, Priyono, and Idrus, 2015) . The assessment conducted by the staff department at STIKes and STMIK Hang Tuah Pekanbaru still uses manual methods with unclear weighting values that cause uncertainty about the results of staff and employee evaluations at STIKES and STMIK Hang Tuah Pekanbaru, thereby affecting the selection of staff and employees who have to get fostered so that more disciplined and who do not need to get fostered. The above problems can be overcome by building a Decision Support System using the Simple Multi Attribute Rating Technique (SMART) method. This method can identify a problem with multiple attributes. 2. LITERATURE REVIEW The concept of Decision Support Systems (DSS) was introduced into the information and computing systems literature by Gorry and Scott Morton in 1971 (Power, Sharda, and Burstein, 2015). DSS simulates the function of human cognitive decision making based on artificial intelligence methodologies (including expert systems, data mining, machine learning, connectionism, logistic reasoning, etc.) (Jao, 2010). Modeling in the construction of DSS is carried out the following steps (Wanto and Damanik, 2015): 1. Feasibility Study (intelligence) In this step, determining objectives and finding procedures, collecting data, identifying problems, classifying problems, until the problem statement is formed 2. Design In this step the model will be formulated to be used and determine the criteria. Then, look for alternative models that can solve the problem. 3. Election (Choice) Rizer Fahlepi… Vol 1(2) 2020 : 103-112 104 In this step, a model selection is carried out, including the solution of the model. After that, sensitivity analysis is carried out by replacing several variables 4. Making DSS After the model is determined, proceed with its implementation into the DSS application. The SMART method is a method used for multi-criteria decision making that was developed by Edward in 1977(Taylor & Love, 2014). This method is based on the theory that each alternative consists of a number of criteria that have values and each of these criteria has a weight that illustrates how important these criteria are compared to other criteria. The grading of these weights is used to assess each alternative in order to get the best alternative (Suryanto and Safrizal, 2015). The SMART method is based on the additive liner model. This means that the overall value of the given alternatives is calculated as the total value of each criterion (attribute) multiplied by the weight of the criterion (Barfod and Leleur, 2013; Bray, 2015). The calculation steps using SMART, namely: 1. Step 1: determine the number of criteria to be used 2. Step 2: provide a scale of 0-100 based on the priorities that have been inputted and then normalized .................................................................... (1) Where: nwj is normalization of criteria weight j wj is the weight value of the jth criterion k is the number of criteria wn is the weight of the nth criterion 3. Step 3: each alternative is given a criterion value 4. Step 4: calculate the utility value for each criterion Calculating utility value: ..................................................... (2) Information: ui (ai): Utility value of criterion 1 for criterion i Cmax: Maximum criteria value Cmin: Minimum criterion value Cout i: The value of the i criteria 5. Step 5: calculate the final value of each .................................... (3) 3. RESEARCH METHOD The stage of the research framework starts from Problem Identification, Data Requirement Analysis, Systems Analysis Using SMART, Design, Implementation of the SMART Method, Testing, and Drawing Conclusions. The framework in this study can be illustrated in Figure 1 below: Rizer Fahlepi… Vol 1(2) 2020 : 103-112 105 Figure 1. Research Methodology 4. RESEARCH RESULTS AND DISCUSSION 1. Criteria Identification At this stage the process of determining what criteria are used in evaluating the performance of staff and employees in the STIKES and STMIK Hang Tuah Pekanbaru. In this study the number of criteria used was as many as 6 criteria for employee appraisal namely Warning, Performance, Attendance, Discipline, Obedience, and Compliance. 2. Alternative Identification At this stage the process of determining alternative alternatives will be carried out. Alternatives in the form of the names of staff and employees at STIKES and STMIK Hang Tuah Pekanbaru namely, Yuda Irawan, S.Kom., M.Kom, Rian Ordila, S.Kom., M.Kom, Leon Chandra, SKM., M.Kes , Jufri, Mardeni, S.Kom., M.Kom. 3. Criteria Weighting Weighting the assessment criteria is given relatively with the provisions of the criteria that have the highest level of importance based on the data of the importance of the criteria in Table 1 given a value of 100 and a minimum value of 10. Rizer Fahlepi… Vol 1(2) 2020 : 103-112 106 Table 1. Criteria Weight No Criteria Weight 1 Commemorative Latter 100 2 Performance 87,5 3 Attendance 70 4 Discipline 62,5 5 Obedience 40 6 Compliance 40 Total 400 4. Criteria Normalization After the criteria weight value is given, the next step is to calculate the normalized value of each criteria weight value by using equation (1): Rizer Fahlepi… Vol 1(2) 2020 : 103-112 107 The results of the calculation of normalization of relative weights can be seen in the following table: Table 2. Normalization Criteria Weights No Criteria Weight Relative Weight (wj) 1 Commemorative Latter 100/400 0,25 2 Performance 87,5/400 0,21875 3 Attendance 70/400 0,175 4 Discipline 62,5/400 0,15625 5 Obedience 40/400 0,1 6 Compliance 40/400 0,1 5. Single Development - Attribute Utilities The next step is to develop single-attribute utilities based on the values given to each alternative. Each criterion is given sub-criteria and the value of the sub-criteria is as seen in table 3. Table 3. Development of Single-Attribute Uttilities No Criteria Sub Criteria Value 1 Commemorative Latter There is No 100 SP 1 85 SP 2 75 SP 3 50 2 Performance Very good 100 Well 85 Pretty good 75 Not good 50 3 Discipline Very good 100 Well 85 Pretty good 75 Not good 50 4 Attendance Very good 100 Well 85 Pretty good 75 Not good 50 5 Obedience Very good 100 Well 85 Pretty good 75 Not good 50 Rizer Fahlepi… Vol 1(2) 2020 : 103-112 108 6 Compliance Very good 100 Well 85 Pretty good 75 Not good 50 After the value of Single-Attribute Utilities is determined, the next step is to determine each criterion value of each staff and employee based on table 3. Table 4. Value of Each Criteria No Employee Name Commemorative Later Performance Attendace Discipline Obedience Compliance 1 Yuda Irawan 100 85 85 85 85 85 2 Rian Ordila 100 85 100 85 85 85 3 Leon Chandra 100 85 85 75 85 85 4 Jufri 75 75 75 50 75 75 5 Mardeni 100 85 100 100 85 85 After the value of each criterion for each employee is determined, then the utilities value calculation process for each employee's criteria is calculated as follows: Commemorative Latter Rizer Fahlepi… Vol 1(2) 2020 : 103-112 109 Performance Attendance Discipline Obedience Rizer Fahlepi… Vol 1(2) 2020 : 103-112 110 Complience After calculating the overall utility values for each criterion, the results are as shown in table 5. Table 5. Alternative Utilities Values No Employee Name Criteria Name Utility Value 1 Yuda Irawan Commemorative letter 100 Performance 70 Attendance 70 Discipline 70 Obedience 70 Compliance 70 2 Rian Ordila Commemorative letter 100 Performance 70 Attendance 100 Discipline 70 Obedience 70 Compliance 70 3 Leon Candra Commemorative letter 100 Performance 70 Attendance 70 Discipline 50 Obedience 70 Compliance 70 4 Jufri Commemorative letter 50 Performance 50 Rizer Fahlepi… Vol 1(2) 2020 : 103-112 111 Attendance 50 Discipline 0 Obedience 50 Compliance 50 5 Mardeni Commemorative letter 100 Performance 70 Attendance 100 Discipline 100 Obedience 70 Compliance 70 6. Calculate End Value After the utility values of each criteria for each staff and employee are obtained, the next step is to calculate the final grade using Eq. (3): Based on the ranking of values obtained, the value can be categorized to: Table 6. Categories of Assessment Results No Value Range Category 1 76-100 Very well 2 50-75 Enough 3 0-49 Do Fostering So from the calculation results obtained by the assessment results as in table 7. Table 7. Employee Performance Assessment Results No Employee Name Rating Result Category 1 Yuda Irawan 77,5 Very Good 2 Rian Ordila 82,75 Very Good Rizer Fahlepi… Vol 1(2) 2020 : 103-112 112 3 Leon Candra 74,375 Moderate 4 Jufri 42.19 Do Coaching 5 Mardeni 87,4375 Very Good 5. CONCLUSIONS From the results of the implementation of the SMART method for identifying employee performance at the STIKES and STMIK Hang Tuah Pekanbaru conclusions can be drawn as follows: 1. The SMART method was successfully implemented into a system developed using the PHP and MySQL programming languages in accordance with the analysis and design created 2. The SMART method can be used to identify employee performance that has 6 criteria. Assessment is carried out by weighting each criterion and producing 3 categories of assessment, namely Very Good, Enough, and Perform Coaching 3. The SMART method can provide more accurate assessment results by weighting each criterion. That way we get more accurate calculation results. 4. From the results of system testing, it is known that the results of the calculation of the SMART method on the developed system are in accordance with the results of manual calculations. Suggestions given by the author for further research are: 1. It is expected that in subsequent studies it can compare the SMART method with several other methods that can be used to identify employee performance. 2. It is expected that in future studies the SMART method can be used in a more complex case analysis or a case that has more criteria used. REFERENCES Barfod, M. B., & Leleur, S. (2013) Multi-criteria decision analysis for use in transport decision making. DTU Lyngby: Technical University of Denmark. Bray, R. (2015). Developing a participative multi criteria decision making technique: a case study. International Journal of Management and Decision Making, 14(1), 66. doi: 10.1504/IJMDM.2015.067381. Jao, C. S. (2010). Decision support systems, Decision Support Systems. inTech. doi: 10.5772/3453. Thaief, I. et al. (2015). Effect of training, compensation and work discipline against employee job performance: (Studies in the office of PT. PLN (Persero) Service Area and Network Malang). Review of European Studies, 7(11), 23–33. doi: 10.5539/res.v7n11p23. Power, D. J., Sharda, R., & Burstein, F. (2015) Decision Support Systems, Wiley Encyclopedia of Management. doi: 10.1002/9781118785317.weom070211. Suryanto & Safrizal, M. (2015). Sistem Pendukung Keputusan Pemilihan Karyawan Teladan dengan Metode SMART (Simple Multi Attribute Rating Technique). Jurnal CoreIT, 1(2), 2460–738. Taylor, J. M. & Love, B. N. (2014). Simple multi-attribute rating technique for renewable energy deployment decisions (SMART REDD). The Journal of Defense Modeling and Simulation: Applications, Methodology, Technology, 11(3), 227–232. doi: 10.1177/1548512914525516. Wanto, A. & Damanik, H. (2015). Analisis Penerapan Sistem Pendukung Keputusan Terhadap Seleksi Penerima Beasiswa BBM ( Bantuan Belajar Mahasiswa ) Pada Perguruan Tinggi Menggunakan Metode Simple Additive Weighting ( SAW ) ( Studi Kasus : AMIK Tunas Bangsa Pematangsiantar). Seminar Nasional Rekayasa (SNTR) II, 323–333.