AD ALTA
JOURNAL OF INTERDISCIPLINARY RESEARCH
Fig.6. The physical data model of the BCPTs data warehouse
The input (Fig. 7) and output screens (Fig. 8) which include the
data that are stored in the target BCPTs data warehouse are
below depicted.
Fig.7. Input BCPTs screen
Fig.8. Ouptut BCPTs screen
The overall BI system architecture is also included for the better
interpretation of the web-based solution (Fig. 9).
Fig. 9.
The proposed BCPTs BI System Architecture
The proposed architecture's main advantage is that the data
loading process is not complicated and not time demanding due
to the avoidance of core ETL (Extraction, Transformation and
Loading) activities such as data formatting, validation and
transformation. The users can directly store the business
continuity data regarding their departments, divisions, functional
areas. Thus, the problem of data extraction from various sources
based on flat files such as excel, text documents, or email
communication, which indicate mandatory ETL activities, is
solved with the existence of the BCPTs API. It can be observed
that the application can be used for exporting data in a
spreadsheet format (.xls or .csv) and utilize it for online
analytical processing operations and data mining predictive
decision making activities.
4.1.3 Online Analytical Processing (OLAP) Operations
The online analytical processing services can be provided via
exporting the data as a spreadsheet document or directly via the
database solution in the form of a query. The proposed schema is
based on the relational database design and implementation
approach. As a consequence, SQL (Structured Query Language)
queries can be utilized as explanatory representative business
continuity OLAP (Sohrabi & Azgomi, 2019) descriptive
operations.
The OLAP covers the analysis services task where the analysis
of the recovery data is based on the UBFRP value for a single
operation. The granularity is based on the operational level
(function, process, activity, and task) as well as the unit level.
Example: The following SQL query is a representative and
simple OLAP aggregate operation. The executed query
computes the average UBFRP for individual business functions
from a determined operational level:
SELECT
BF_RecoveryDATA.BusinesOperation_ID,
Avg(BF_RecoveryDATA.UBFRP) AS AvgOfUBFRP
FROM
BusinessOperationLevel
INNER JOIN
BF_RecoveryDATA
ON
BusinessOperationLevel.BusinessOperation_ID
= BF_RecoveryDATA.BusinessOperation_ID
GROUP BY
BF_RecoveryDATA.BusinessOperation_ID;
4.2 Safety Critical Computations Based on the Exported
BCPTs Spreadsheet Data: Evidence from Real University
BCM Data
4.2.1 Validation of the BCPTs speedy classifier and the
computations supported by the proposed BI tool
The currently proposed contribution, as every proposed business
intelligence solution, serves as a tool for effective and efficient
predictive decision making. Predictive analytics that facilitate
crucial decisions regarding future trends in public organizations
are based on machine learning activities. The current version of
the proposed solution supports classification of critical business
operations based on the UBFRP input. As a consequence,
according to Rule 1 (Section 3) the UBFRP input recovery
variable can indicate the Maximum Recovery Time Effort
(RTE
MAX
) required to recover a business operation.
The currently developed web-based business intelligence
application (Fig.5, Fig.6) has been used to validate the business
continuity management policies based on the BCPTs
computations. Taking into consideration real BCM data from a
public university (Columbus Technical College, 2018) a
spreadsheet dataset in .csv format has been created. Business
continuity parameters have been used as the input data to infer
robust maximum recovery time computations via the proposed
business intelligence schema. The data have been stored in the
physical database from which they have been exported in the
form of a .csv file. Part of the data set is below depicted (Tab.1).
The table includes arbitrary data related to 7 selected safety-
critical operations in a public university. The full data set can be
accessed via the link:
- 361 -