The Benefits of Evaluating the Jordanian Social Security ... · THE BENEFITS OF EVALUATING THE JORDANIAN SOCIAL SECURITY USING DECISION MODELING AND NOTATION Ahmad Zyad Alghzawi,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Thus business rules are central to decision management. Decision management maximizes the ROI of
these business rules by applying them effectivity. Similarly, data mining and predictive analytics provide
insight management (Debevoise, Taylor et al. 2014). Decision management multiplies the value of
analytics to the business result by focusing mainly on numerous repetitive day-to-day decisions
(Debevoise, Taylor et al. 2014, Hilmi, Safa et al. 2017).
From the perspective of BPM, decision management achieves three end goals (Debevoise, Taylor et al.
2014) :
1. Identifying decisions within business processes, whether these are automated or manual;
2. Precisely and unambiguously representing and populating a decision model to specify how the decision should be made who should do it without adding this information to the process model itself; and
3. Implementing, reporting and updating processes in order to continually refine the effectiveness of decision making as well as improving the efficiency of the processes.
Business Process Modeling and Notation (BPMN)
The primary goal of BPMN is to provide a notation that is readily understandable by business users,
ranging from the business analysts who sketch the initial drafts of the processes to the technical
developers responsible for actually implementing them, and finally to the business staff deploying and
monitoring such processes. BPMN was originally published in 2004 by the Business Process Modeling
Initiative as a graphical notation (partially inspired by UML Activity Diagrams) to represent the graphical
layout of business processes. The ever increasing number of adoptions by companies and the growing
interest for this notation caused the adoption of BPMN as Objective Management Group (OMG)
standard in 2006 (Chinosi and Trombetta 2012, von Rosing, White et al. 2015).
BPMN provides a graphical notation in order to represent a business process as a Business Process
Diagram (BPD). The development of a BPMN is based on a flowcharting technique tailored to creating
graphical models of business process operations. A Business Process Model, then, is a network of
graphical objects, which are active ties (i.e., work) and the flow controls that define their order of
performance. Activities are represented as rectangles and diamonds represent alternative workflow
paths (von Rosing, White et al. 2015).
It should be emphasized that one of the drivers for the development of BPMN is to create a simple
mechanism for creating business process models, while at the same time being able to handle the
complexity inherent to business processes (A.white July, 2004). The approach taken to handle these
two conflicting requirements was to organize the graphical aspects of the notation into specific
categories. This provides a small set of notation categories so that the reader of a BPD can easily
recognize the basic types of elements and understand the diagram. Within the basic categories of
elements, additional variation and information can be added to support the requirements for complexity
without dramatically changing the basic look-and-feel of the diagram (Aguilar-Savén 2004, Gabryelczyk
and Jurczuk 2017).
BPMN has several uses. It is used to communicate a wide variety of information to different audiences.
BPMN is designed to cover many types of modeling and allows the creation of process segments as
well as end-to-end business processes, at different levels of fidelity. Within the variety of process
modeling objectives, there are two basic types of models that can be created with a BPD (Gabryelczyk
and Jurczuk 2017):
• Collaborative (Public) B2B Processes and
• Internal (Private) Business Processes
Decision Model and Notation:
In the last few years, the OMG (Object Management Group) proposed a new innovation, which is called
“Decision Model and Notation” (DMN) specification. In February 2014, version 1.0 of the DMN
specification was approved by OMG’s board (OMG: ). Since March 2016, version 1.2 has been
available via their website.
DMN Overview
DMN stands for Decision Model and Notation. The essential goals of DMN is to support a common
notation for decision login that is understandable for business users, business analysts and developers
alike. DMN supplies the constructs for the decision-making process itself and allows to model decision
rules.
A DMN model defines two levels: the decision requirements graph (DRG) and the decision logic. Where
the required information is coming from and how it can be depicted in one or more decision
requirements diagrams (DRDs) is the object of the former, whereas the latter describes the logic behind
the decision, depicted in the Decision Table [15]. The upper half of a decision table indicates all
potential combinations of conditions that may lead to certain actions, while the bottom half describes the
actions to be taken (i.e., outcomes). A minimal scope is specified for the standardization by OMG
because the goal of DMN is to offer support to other decision logic notations (e.g., decision trees) and to
allow for references to other types of models (e.g., SBVR).
Connecting between DMN and BPMN :
Decision Management aims to separate decision logic from the process logic. Large sequences of
gateways and checks, signalling decision logic, will all be removed from the BPMN model and captured
within a single decision step. A decision-driven process then uses the outcome of the evaluation of this
decision logic in several possible ways, including (Debevoise, Taylor et al. 2014):
Changing the sequence of activities that are taken after a decision, including what the next activity or process that is required to meet the directive of the process.
When analyzing the efficiency of the proposed process models, some apparent characteristics are easy
to derive from the process models. Firstly, there are a lot of steps that involve data extraction or data
validation. It triggers the question why those validations are executed in such a fragmented sequence.
Do some of these steps represent double work or, if not, could they also be merged into a single
activity? A deeper analysis on task-level rather than process-level could reveal some opportunities to
make the process more efficient.
Furthermore, as can be seen in the Table below the subprocesses seem to have some common
activities; It would be beneficial to examine to what extend these tasks really are similar to each other
and whether the four different processes could actually be collapsed into a single model.
Natural death Partial disability Work-related RetirementCollect signature of heirs X Attach documents X Extract data X X X X Verify entitlement X X X X Verify reliability X X X X Verify intervention periods X X X X Verify employment injury X X X Confirm early retirement X Confirm former one-time payment X Confirm former unemployment allowance
X
The BPMN for the calculation of the pension is mentioned in (Figure 10) (Security 2014):
reduction. Reduction is decided based on the sub-decisions average salary, also age and gender which
are both considered as input data. The second sub-decision is average salary based on the salary
which is considered as input data. The Final sub-decision is the dependent based on the sub-decisions
average salary and the number of dependent which the retirements have and the minima is (0) and the
maximum is (3) (office 2015).
Overview of decisions and sub-decisions (Security 2014)
Decision Question Answer Pension Calculate the pension salary that will be paid out to the
applicant on a monthly basis (Non-negative) Monetary value in JD.
Average salary
Calculate the average salary according to the last 2 years salary the retirements had
(Non-negative) Monetary value in JD.
Dependent how many dependent the retirements have (Non-negative) Number. Redaction How much discount the JSS will do according to the law (Non-negative) Monetary
value in JD. Table 2 explains the pension decision (office 2015).
Pension salary =Average of salary + Dependent-Redaction + 20
Table 2 explains the main decisions shown in the model above (Fig 12). We can see the calculation of
the pension according to this type of decision from the table. The sub-decisions average salary, number
of dependent people and conditions for reduction, which we will explain later in this section. The input
data are gender and years of contribution.
In the left column of the table, three different pension decisions can be found. Also the eligibility,
average salary, dependent and redaction are sub-decision and gender, years of contribution input data
to have the output pension salary.
In our case for the calculation of the DMN of the Current and new Early pension, the decision has more
than one sub-decision In Table 3. we will explain the first sub-decision, namely the average salary
(office 2015).
Table 3: Average salary sub-decision
F Salary (Number) Average Salary (Currency) 1 ≤ 50 JD 50 JD 2 ≤ 1500 JD = 0.025 * Years of contribution * Salary 3 > 1500 JD = 0.025 * Years of contribution * 1500 + Years of contribution * (Salary – 1500)
In this sub-decision as we can see that there are only three conditions, namely the salary before
retirement and the years of contribution. If the salary is either (≤50) or (≤1500) or (≥1500) and years of
contribution are between 15 and 30, the JSS does the average of the pension calculation as we can see
in table (3).
In this Table(4) we will explain the second sub-decision, namely the number of Dependent people (office
Number Number 1 = 0 0 2 = 1 = Average Salary * 0.12 3 = 2 = Average Salary * (0.02 + 0.06) 4 ≥ 3 = Average Salary * (0.02 + 0.06 + 0.06)
As we mentioned before and as we can see from Table (4) in the JSS the minimum number of
dependent people is (0) and the maximum is (3). In this case there is only one input data number of
dependent people and based on the other sub-decision, the average salary, retiring people will
according to the number of dependent people get extra’s on their pension.
In some cases, the retirements get some reduction on their pension and in Table (5) we will present the
cases. That is the last sub-decision for the main pension decision (office 2015).
Table 5: Reduction sub-decision
F Age [45-∞] Gender {Male,Female} Reduction Currency(JD) 1 <46 Male =Average of salary*0.2 2 <47 Male =Average of salary*0.18 3 <48 Male =Average of salary*0.16 4 <49 Male =Average of salary*0.14 5 <50 Male =Average of salary*0.12 6 <51 Male =Average of salary*0.11 7 <52 Male =Average of salary*0.1 8 <53 Male =Average of salary*0.09 9 <54 Male =Average of salary*0.08 10 <55 Male =Average of salary*0.07 11 <56 Male =Average of salary*0.06 12 <57 Male =Average of salary*0.05 13 <58 Male =Average of salary*0.04 14 <59 Male =Average of salary*0.03 15 <60 Male =Average of salary*0.02 16 ≥60 Male 0 17 <46 Female =Average of salary*0.14 18 <47 Female =Average of salary*0.12 19 <48 Female =Average of salary*0.1 20 <49 Female =Average of salary*0.08 21 <50 Female =Average of salary*0.07 22 <51 Female =Average of salary*0.06 23 <52 Female =Average of salary*0.05 24 <53 Female =Average of salary*0.04 25 <54 Female =Average of salary*0.03 26 <55 Female =Average of salary*0.02 27 ≥55 Female 0
In this sub-decision we can see that there are three conditions. Two of them are the input data age and
gender and the third sub-decision is the average of salary. The table seems to be designed to
In this sub-decision as we can see that there are four input data, his or her actual age, gender, namely
the salary and the years of contribution. the JSS does the average of the pension calculation in this
type of decision as we can see in Fig (14).
In this table (9) we will explain the third sub-decision, namely the number of Dependent people (office
2015).
Table 9: Normal Old Age Dependent sub-decision U Number of dependents Dependent Number Currency 1 = 0 0 JD 2 = 1 = Average of salary * 0.12 3 = 2 = Average of salary * (0.12 + 0.06) 4 ≥ 3 = Average of salary * (0.12 + 0.06 + 0.06)
As we mentioned before and as we can see from Fig (19) in the JSS the minimum number of dependent
people is (0) and the maximum is (3). In this case there is only one input data number of dependent
people and based on the other sub-decision, the average salary, retiring people will according to the
number of dependent people get extra’s on their pension.
This DMN of the OLD Pension Eligibility (Figure 15) (office 2015):
Fig.15 The OLD pension Eligibility
Overview of OLD Eligibility decisions(office 2015).
decision Question Answer Eligibility If the retirements can have the pension according to JSS condition Yes or No The main decision in the model above (Fig 15)is OLD Pension Eligibility. OLD Pension is decided based
on the input data age, gender and year of contribution.
In table (10) we explain the Eligibility of one person for a pension according to different decision
Using these DMN models, we could make a simulation to support an hypothetical 45-year old man with
a salary of 600JD in making his retirement decision. Table (18) compares four different situations where
he would either retire immediately or wait until 50, 55 or 60 years of age. Based on age and years of
contribution. In general, if the retired go to the pension he will receive less pension early per month put
he will get more money from JSS in total. However, the applicant should consider that in case of early
retirement, he loses his current salary of 600JD, which amounts to 7200JD annually. Disregarding
factors such as stress or physical complaints, it seems that in the current system, the lost income would
be the main motivation for people to apply for a pension at a later stage.
Table 18: Different retirement scenarios for a 45 year old man, having currently 3 dependents and a
salary of 600JD.
Age of retirement 45years 50 years 55 years 60 years Years of contribution 25 30 30 30 Years having 3 dep 6 years 1 0 0 Yeas having 2 dep 10 years 10 years 5 years 0 Years Having 1 dep 15 years 15 years 15 years 15 years First pension 438.95 568.19 591.14 544 Total pension until 75 154289.4 140864.28 121364.4 97920 Average Yearly Pension 4977 5418 5779 6528 Salary earned from employer (since 45) 0 36000 72000 108000
Given the logic defined in DMN, we could make projections for an entire population if we have sufficient
information about the demographics. Table 5 displays a simplistic example of demographics in different
scenarios. Based on current numbers, 12188 new people go on retirement. In general, 61% of them
enjoy early retirement around the age of 50. The other 39% receives old-age retirement at an average
age of 60. Best case scenario is considered an optimistic 10% early retirees, the remainder receiving
old-age pension. In a worst case, one could see as much as 80% of applicants for early retirement.
Based on these estimated ratios, we can calculate the total number of applicants in each scenario. It
can be seen that the total number of applicants is lowest when the fewest people apply for early