Fraud Analytics: Challenges to Overcome

September 7, 2016

Organizations typically limit their fraud data analysis to quantifying the financial impact when fraud is detected by some other means. However, data analytics techniques can also play a significant role in early-warning, detection and monitoring of fraud.

 

Fraud analytics

 

While analytics offers many advantages as a prevention measure, it is critical that firms recognize the challenges associating with performing these techniques:

  • Quality of the Data– The results from analytics tests can only be as good as the data you feed in as your input. Before performing tests, it is important to assess the quality of data and perform validation and cleansing as required. Garbage in, garbage out, as they say.
  • Data volumes – Organizations are awash with data from multiple touch points and there may be significant data volumes supporting certain business processes. Your data analytics testing infrastructure should be robust enough such that it can handle these huge volumes.
  • Data security – It is essential that security protocols are considered throughout the extraction and analysis protect the confidentiality and integrity of the source data.
  • Skillset gap – Data analytics requires a combination of business acumen (holistic understanding of business requirements) and technical skills (know-how of tools and techniques) to define the tests, perform the analysis and interpret results in order to generate meaningful insights. Very few people possess both business and technical skills, and this lack of human capital to perform quality analyses is a serious challenge that organizations need to overcome through rigorous candidate selection and on the job training.

Imarticus is hosting an Executive Development Program on Big Data & Credit Risk Management. More details to be revealed shortly!

Post a comment

3 × 5 =