|
by Jotham Mapundi, Director Public Sector at SAS Institute
A major challenge for public sector bodies is not just the quality of data, but the mere presence of it. That said the data that is at the hands of the government organisations in South Africa is, in many instances, incomplete and inconsistent, and often out of date. It is difficult to perform true analytics on data where, for example, 5% of the people in the database are deceased, and another large percentage has moved.
While the nirvana for public sector departments the world over is to have a single view of its citizens, the reality is a far cry from that fact, as access to data which encompasses all of the above-mentioned aspects is possibly more challenging to them than their corporate counterparts.
There are four major challenges facing the collection and integrity of data in the public sector, which cause problems when trying to pull intelligence out of the available information, or when trying to run BI solutions on this data. These include accuracy, inconsistency, timeliness and availability issues.
Accuracy issues
Problems of accuracy occur when stated facts do not match reality. This can be because data is often times out of date, corrupt or has been incorrectly captured. Traditionally government agencies have collected data via forms, which can often mean that data captured is in fact incomplete or inaccurate as the right information is not revealed.
Sometimes forms are also used across purposes to save money, making some of the information redundant or even irrelevant, and when irrelevant gaps in forms are left blank, it opens the way for incompleteness. Blank fields when dealing with data analysis are bad, as they can often times represent nil or be taken to mean zero, if analytical tools are not intelligent enough to see past the values.
Inconsistency issues
Data which is inconsistent or inaccurate can cause real problems when trying to draw facts from it. For example, Gauteng abbreviated as GP, GTG or Gaut, may cause problems in a system that has not been coded to make room for exceptions. The knock on effect could leave several departments with the wrong information as opposed to merely a single instance.
When working with business systems, with data inconsistencies one may have problems when sending out bills and other information to customers. In the case of a municipality an incorrect form may create a duplicate in the system and in error send out two bills to the same person.
Timeliness issues
Data that is out of date can unnecessarily load your system and clog your reports, while at the same time hamper the accuracy of the reports you are trying to build. Government departments have to sometimes rely on data that is as old as birth data, as no other contact has been made with the citizen since. Sometimes rather than not crunching the information, old data or estimates are used.
Data availability
Sometimes the way in which data has been collected is just wrong, or incompatible with your existing workflow environment. When data is not immediately available it can leave flaws in your intelligence, or gaps in the decision making process. Sometimes the data has just never been collected and therefore is just not available.
Getting it right
There are so many good business cases for BI in the public sector. A single view of citizens, being able to measure socio-economic factors and their impact on people development, ensuring funding to the right areas, consolidation of citizen information and access to this - we really do not need to justify the need for it.
The reality however is that data quality issues scare the public sector 'enterprises' off in many instances. But one needs to start somewhere and there are a number of tools available today that do not need you to reinvent the wheel, they simply need for you to have the business case to ensure data integrity. To quote the age old IT expression ... Garbage In, Garbage Out.
|
Features 
Press Releases
