AD ALTA
JOURNAL OF INTERDISCIPLINARY RESEARCH
UBFRP = (1*1.5+1*1)+ (2*1.5+2*1+1*0.5)+(8*1.5)=20 points.
3.2 Adjusted Business Function Recovery Points (ABFRP)
For the computation of the specific value the following recovery
complexity paremeters are considered:
Technical Recovery Factors (TRF)
Environmental Recovery Factors (ERF)
Unexpected Recovery Factors (URF), and
Recovery Time Effort (RTE)
*
**
ABFRP
TRFERFURFUBFRP
=
(2)
2
5000
3
RTE
ABFRP
=
−
(3)
The Business Continuity Points method relies on the recovery
complexity concept following the software/system complexity
principle of Karner (1993) to estimate the effort required to
develop an information system.
According to lab computations, the criticality ranking can be,
however, determined without the computation of the resumption
timeframes for specific UBFRP values (Podaras et al., 2016).
The current work presents the developed decision support data
warehouse schema, which currently implements the criticality
ranking of individual processes without the computation of the
resumption timeframes. The specific classification is entitled
"speedy" criticality ranking. The more detailed criticality
ranking data warehouse solution is currently under development.
Based on preliminary lab computations and after validating the
general BCPTs business rules (Podaras, 2018) the following
decision making algorithm has been generated and applied in the
current BI solution.
Rule 1: “speedy classification of a business function based on
UBFRP”
Empirical lab computations led to the construction of a data set
including 46 business functions which has been used for
machine learning classification of a business function based on
UBFRP input (Podaras, 2018). The classification rule induced
via this study is the following:
IF UBFRP<9.7 Points THEN
IF UBFRP<14.45 Points THEN
IF UBFRP<20.89 Points THEN
Criticality Level = L2
(Critical Operation
RTE
MAX
ELSE Criticality Level
= L1
Critical
Operation
RTE
=24Hours)
MAX
END IF
=2Hours)
Criticality Level = L3
(Non-Critical Operation
RTE
MAX
ELSE Criticality Level = L2
(Critical Operation RTE
=72hours)
MAX
ENDIF
=24Hours)
Criticality Level =
L4 (Non-Critical
Operation RTE
MAX
ELSE Criticality Level = L3 (Non-Critical
Operation RTE
=168Hours)
MAX
END IF
=72Hours)
Based on this rule, a speedy classification can be implemented
with approximately high accuracy. However, only the maximum
recovery time that is mapped to the corresponding IVL (Gibson,
2010) can be assigned. Precise recovery timeframes cannot be
determined via the speedy classifier.
Rule 2: The Recovery Scenario (RS) Selection
The rule-based Recovery Scenario (RS) selection of individual
operations is illustrated as a decision tree which has been derived
via the R software package (Yadav & Roychoudhury, 2018)
after importing and processing the lab-based empirically derived
data (Fig. 2)
Fig. 2. The decision tree for selecting the appropriate recovery
scenario. Source: (Podaras, 2018).
The semantics used in the above illustrated decision tree has the
following meaning:
Simple RS: TRF=URF=ERF=0.85. The value is constant
which means no international units are utilized (Podaras et
al., 2016)
Average RS: TRF=URF=ERF=1 and
Complex RS: TRF=URF=ERF=1.15.
As it was previously stated, the scenario selection and the
computation of these parameters is important for the precise
computation of the recovery time which is not included in the
current version of the BI solution.
3.2 Business Intelligence Data Warehouse Preliminaries
Multiple academic researchers and business experts have
provided precise delineation and definition with respect to the
business intelligence data warehouse systems. A representative
definition considers a data warehouse as “a collection of
methods, techniques, and tools used to support knowledge
workers — e.g., senior managers, directors, etc. — to conduct
data analysis that helps with performing decision making
processes and improving information resources” (Golfarelli &
Rizzi, 2009). When data warehouse systems are integrated, a
standard procedure regarding the design process is the
consideration of multiple dimensions, the facts which indicate
the measurable variables of these dimensions and the key
attributes for dimensions and facts (Romero & Abelló, 2010).
The data warehouse schema consists of several dimensions and a
single fact is known as multidimensional schema or star schema.
In the multidimensional schema, "facts correspond to events
which are usually associated with numeric values known as
measures and are referenced using the dimension elements”
(Caniupán et al., 2012). Moreover, “dimensions are modelled as
hierarchies of elements, where each element belongs to a
category. The categories are also organized into a hierarchy
called hierarchy schema.” (Caniupán et al., 2012).
Finally, based on the traditional design approaches regarding the
relational as well as the object-oriented database models, three
relevant design categories are distinguished, that is the
conceptual, logical and physical design (Vaisman & Zimanyi,
2014). In the results section both the conceptual and the physical
design of the proposed data warehouse are illustrated due to their
importance.
- 359 -