The South African Revenue Services (SARS) is undergoing an extensive modernisation process that will greatly enhance its enforcement capabilities, according to SARS manager of data analytics and mining, Eugene Wessels.
Wessels was speaking on the second day of ITWeb's annual Business Intelligence Summit and Awards in Bryanston.
SARS has moved from a system where data was given to business departments straight from the source to one where its information management department enables business users to manage their own BI capabilities.
Summaries and reports with different analytics are given to different levels of business based on their needs: The executive-level users are given strategic summaries with KPIs, regional offices are given tactical reports, and at the branch level users are given operational intelligence.
Wessels said SARS has learnt it is important for business and IT to work together and understand each other when it comes to BI.
He went on to demonstrate that there is a difference between public and private sector BI use where the private sector makes use of it for products, profits, and chasing competitive advantage, the public sector uses it for deciding on policy, enforcement and compliance measurement.
SARS is watching
As an example, Wessels showed how SARS would use BI to go about finding a fraudulent company or person, at the same time demonstrating the system it currently has in place.
He used the instance of a company that under-reports its corporate income tax and over-reports the VAT it pays, so it receives continuous refunds. Using advanced analytics, SARS could predict values in non-mandatory fields that were not filled in and statistically infer outliers. SARS could then use text-mining of suspicious activity reports to find patterns, which could be fed into a risk engine.
Analytics could also be used to predict which of these suspicious parties would be most likely to be reached by a call centre, and to predict which are most likely to receive a successful audit. SARS is then able to prioritise its work and even use associative and geospacial data to find clusters of and links between the suspicious parties.
SARS was the runner-up for the ITWeb Business Intelligence Excellence Award 2010.
Share