Finance seems to be an ideal venue for advanced business intelligence methods. The field moves at a high speed and is based on managing and monitoring transactional information. Processes can accomplish everything from fraud detection to simple transactional agility building. According to Computerworld contributor Susan Feinberg, the financial industry's BI users should focus on high speed projects and exciting new processes as agile new analytics tools become available.
New data types
Big data has changed many industries, including finance. While banks have long had access to transactional numbers as an analytics source, they can now introduce context to processes through other types of data. Feinberg singled out three areas central to analytics in the financial industry, some structured and some not. She noted that these areas should be carefully monitored, with heavy governance attention to ensure processes run smoothly.
Feinberg stated that banks should keep a close watch on the classic transactional data that makes up basic account management and complement it with reporting information meant for customers and government regulators. This data often comes from many different sections within a company. For this reason, it resembles big data in its usage cases even when the actual amount of information is comparatively small.
Fraud detection processes represent another advanced data function banks can undertaken with the aid of advanced new analytics programs. Feinberg singled them out as a big data project in the financial industry that has surged ahead, whereas most unstructured information plans have not yet become mission critical.
The actual mechanics of using big data in finance involve combining information sources to gain a composite view of a problem. Sourcing Focus contributor Simon Asplen-Taylor recently explained how firms fight fraud through analytics. Algorithms have existed for years to detect odd and possibly criminal activity in accounts. The big data breakthrough involves using unstructured customer data as context.
Asplen-Taylor noted that, armed with data from online sources like social media, banks can determine the actual likelihood that the fraud detection algorithm has made a false positive. Cutting down on these mistaken red flags can save banks significant money by cancelling wrongful investigations in the early stages.
To be effective, fraud detection and the systems surrounding it must be fast, nearly real-time. This is where big data systems excel. Though information reserves are bigger than ever, sorting through them has gotten faster, rather than slower.