AHP-TOPSIS for analyzing job performance with factor evaluation system and process mining

Job performance is a type of assessments which refers to scalable actions, behaviour and outcomes that employees engage in or bring out linked with and contribute to organizational goals.This research employed the Factor Evaluation System (FES) method to analyze the job performance due to the common usage of the method. In analyzing employees, FES consists of nine factors; however, those nine factors are considered to be insufficient. Hence, the researchers used the process mining method to improve FES. Process mining analyzes job performance in details. The steps taken in process mining consist of time stamp, case, activity, and resources of employee. This means that the method can be continuously used, since the researcher provides weight for each factor. The weight of each factor is obtained from Analytic Hierarchy Process-Technique for Order Preference by Similarity to Ideal Solution. The result shows that FES with process mining are good for job performance but AHP-TOPSIS is considered to be incompatible for usage compared to the real work because the priority of the FES factors from the method is inconsistent with the priority factor made by manager of the warehouse officer.


Introduction
Job performance is one type of assessment which refers to scalable actions, behavior and outcomes that employees engage in or bring out linked with and contribute to organizational goals [1][2][3][4].Contextual performance and task performance constitute types of employee behavior.Contextual performance focuses on individual performance while task performance focuses on core technical skill [5].
Many researchers used the Factor Evaluation System method to analyze job performance, so did this paper [6,7].Factor Evaluation System is a commonly used method in determining levels within an organization.Each level shows the degree of difficulty and responsibility of a position which, at the same time, indicates the level of wage and level of qualification required in that position [7].FES uses nine factors to analyze the position.There are 2 types of FES; standard FES to classify functional position and standard FES alongside the classification to manager to evaluate structural positions.In previous study, FES usually was only used to help the company to obtain the position performance evaluation that influence work achievement which has to be connected with position burden.
FES should be available for review, since the document is public.Nevertheless, FES is unable to deal with changes because FES has been predetermined regarding the nine factor rules.Currently, we need to analyze every activity to obtain the best result for performance evaluation.To improve the analysis result using FES, this research combines aforementioned method with the process mining [8].Until recently, a study discussing the FES method with business processes remains a rarity.
Process mining is a technique in terms of process management which supports the analysis of business processes based on event logs [9][10][11].It is a set of techniques which automatically construct a model of an organization's current activities and its major activities variations [12].To extract the information from the real executions constitutes one of the objectives of process mining [13,14].From the event log of company, the history of time stamp, case, activity, and resources of employee can be identified [15,16].Besides containing activity TELKOMNIKA ISSN: 1693-6930  AHP-TOPSIS for analyzing job performance with factor evaluation system... (Gabriel Sophia) 1345 and case, event log also contains information about trace [17].Process mining assists the FES method to further analyze the factors [18].In order that this method not only used to measure the employee performance, weight for each factor must be identified.In the previous study, researcher use AHP to obtain the weight for the job performance.Thus, this paper uses another method namely AHP-TOPSIS (Analytical Hierarchy Process Technique for Order Preference by Similarity to Ideal Solution) to obtain the weight and order [19][20][21].TOPSIS is used to obtain more accurate information regarding the weight.In this research report, the researcher proposed a notion to analyze job performance using combination of FES with process mining and AHP-TOPSIS.

Research Method
The objectives of this research is to recalculate using this method.The steps are as follows.

Step 1: Identifying the Standard Operating Procedure (SOP) from the Company
An example of SOP is shown in Figure 1, while Table 1 presents the attributes of the events in the logs.The table consist of Case ID, Activity, Start and Finish Time and Resources showing the activity doer.YAWL (Yet Another Workflow Language) is an application used to create a business process model.Using YAWL, detail activity can be identified because the result of YAWL is based on record logs.The initial SOP in Microsoft Visio format was redrawn using the YAWL format to later be simulated.Drawing the SOP in YAWL format was not only based on the initial SOP, but also on the event log. Figure 2 is the main SOP or SOP level 1 of Receipt Items.

Figure 2. SOP level 1-receipt items
The result of the redrawn SOP is possibly the same as the initial, but most result are usually longer.In Figure 2, three double box buttons are present indicating the fact that certain activity can be more in detailis and is presented in next figure.Figure 3 shows the continuation activity from the SOP of Receipt Items that usually called SOP level 2.  After the business process model was created, the next procedure was to categorize every activity undertaken by the warehouse officer position.Each activity was analyzed using the nine FES factors.Table 2 presents the collection of the activity from SOP Received Items Level 1 and Level 2. Other than that, all activities in SOP level 1 and level 2 have resources and division in charge in that.The result initially consisted of 10 activities.In that  2, 19 activities from one SOP are obtained.There are 3 activities (confirming arrival items by email, sending packing list of arrival item and writing the seal number) which as a matter of fact were done by another position, i.e the purchasing and admin sales.Consequently, only 16 activities are included for the warehouse officer position in SOP Received Items.Table 3 shows the value of each FES factor (column 2-9) to the activity (column 1).

Step 4: Every Position Consists of Different Factor which Has More Influence from the Other
Its is required to determine the hierarchy from the highest to the lowest factor by asking the manager.After obtaining the priority factor, the FES was calculated using AHP.The basic step in the AHP method are [22,23]: 1) Structure elements in criteria, sub-criteria, alternatives, etc 2) Make a pair-wise comparison of elements in each group 3) Calculating the weighting 4) Calculating the Consistency Index (CI) To calculate the weight with AHP the list of from Table 4 was used.Table 4 shows the the weight of FES factor from the AHP which will be used in the next step.To test the weight consistency of AHP, so the next step is calculated the value of CI and CR, of which CR value<10%.

Step 5: Proceeding the Calculation Using TOPSIS
TOPSIS is a method introduced by Yoon and Hwang [24,25], in which alternative being the closest distance to a positive ideal solution and the farthest distance from the negative ideal solution are chosen.The steps are shown below.1) Normalization of decision matrix.Normalization with TOPSIS starts by squaring all factors followed by dividing the initial value by the total squares.Table 5 shows the normalized each factor with TOPSIS.
2) Weighting on a normalized matrix.Weight is obtained from the AHP result.Table 6 is the result of the matrix multiplication between weight and normalization factor.3) Determining the ideal positive solution and the ideal negative solution.The positive ideal solution is denoted by A+ and the negative ideal solution is denoted by A-.
where: = matrix element row i and coloum j J = {j = 1,2,3,…n and j associated with benefit criteria} J' = {j = 1,2,3,…n and j associated with cost criteria} From Table 6, the minimum and maximum score for each factor of FES is obtained.The next procedure is determining the ideal positive solution and the ideal negative solution.Table 7 shows the result of calculation score from the matrix which consist of positive ideal solution (A+) and negative ideal solution (A-).The proximity relative to positivef ideal in AHP-TOPSIS denote weight for factor FES.The highest score indicates that one factor is more important than the others.Table 9 shows the sequence of weights to the FES factors.

Step 6: Comparing Average Value
The last step is comparing the average value of each activity with the weight factor.

Results and Analysis
From the analysis, not only is the value from every activity, but also the best score.The best score is used to determine the new standard in the recruitment process of the position.Moveover, using the FES method and process mining in the event log for a period time enables to analyze the average and total time of certain task.It can be used by the company to identify  ISSN: 1693-6930 TELKOMNIKA Vol.17, No. 3, June 2019: 1344-1351 1350 the capability of employee in order to find out which one needs training or reward.To obtain such data, researchers can use a method called AHP-TOPSIS to obtain the weight and order of the priority.Upon the completion of the AHP analysis, the result was used for the TOPSIS analysis.
Table 10 shows the sequence of the FES factors from the highest to the lowest.From the table, it is obtained that the most influential factor from FES is purpose of the contacts factor based on AHP-TOPSIS.However, this factor is considered to be inconsistent with the factor priority made by manager of the warehouse officer as shown in Table 5. Job performance combined with process mining generates a more detailed analysis about employee assessment because it enables to elaborate every activity that has been done by the worker.The AHP-TOPSIS completes the analysis by producing not only the weight of each factor but also giving an important sequence of FES factors.

Conclusion
Based on the research, this report proposed a hybrid method in analyzing job performance using Factor Evaluation System and process mining.Process mining helps the FES method to obtain in-details information regarding the activities.The result shows that the initial 10 activities from initial SOP become 19 activities in the latter analysis.It means that the FES method is able to analyze all activities in a position in addition to analyzing the position in general.
In order to use the result of analysis in long term and as a reference in recruiting new employee, this paper report uses AHP-TOPSIS to obtain the weight of each factor.The result shows that AHP-TOPSIS can give the weight and sequence of FES factors that can used to standard for each activity but the values still are incompatible for use compared to the real work because the result obtained using AHP-TOPSIS is not equivalent to the sequence predetermined by the manager.For future research, it is considered to be imperative to conduct in-detail analysis using another method in order to obtain the weight combination for the FES method and process mining.

Table 1 .
Example of Event Log

Table 3 .
Activity with Factor FES in SOP Level 1 & Level 2

Table 4 .
Weights of FES Factor from AHP

Table 7 .
Calculation of Positive Ideal Solution and Negative Ideal Solution