InformationWeek.com

Welcome Guest. | Log In| Register | Membership Benefits

March 27, 2000

Business Intelligence Pays Dividends
Companies and governments use software to reduce losses due to fraud

By Talia Baron

Illustration by Hogan/Braun

Related links:

Telcos Turn To Analytical Tools To Stay In Touch (3/13/00)

The Drive To Analyze (2/14/00)

TechEncyclopedia

Need a definition of a technology term? Look it up here:

Send Us Your Feedback

Heller says the new system lets users place a fraud event in the context of other activities. "Let's say we have a fraud event in relation to a batch of stolen cards," he says. "This system lets us slice and dice the data by looking at which merchants and which countries are involved."

Although Visa has always analyzed such information, the MicroStrategy software lets it perform the analysis much faster. "It also gives us one drill path so that we no longer have to ask the system 10 separate queries and wait for 10 separate responses," Heller says. Visa can find out who was involved, what merchandise was targeted, when it happened, where it happened, and so on--and it can do all of that at once.

Fraud-detection systems are the lifeblood of organizations like NASD Regulations Inc. The independent subsidiary of the National Association of Securities Dealers Inc. is charged with regulating the securities industry and the Nasdaq stock exchange, which in 1999 alone meant monitoring an average of 1 billion shares--and 1 million trades--per day.

To simplify its operations, NASD Regulations implemented the Advanced Detection System in 1997. The internally developed market-surveillance, data mining, and fraud-detection application monitors all Nasdaq transactions for late trade reporting and trade-execution violations.

The system gathers comprehensive market-surveillance data and helps users catch a greater number of fraudulent activities, says Steve Luparello, senior VP of the market regulations division with NASD Regulations, in Rockville, Md. While NASD Regulations has had customized fraud- detection systems in place since the 1970s, the Advanced Detection System has helped simplify the work of market analysts who review the data. "It's given them more in-depth knowledge, more evidence than they've had before," Luparello says. "The system looks for patterns in data on a daily basis and searches for scenarios that may be evidence of fraudulent activity."

Steve Luparello Photo by Gary Parker Making market analysts more efficient was also the goal at Chase Securities Inc., a wholly owned subsidiary of Chase Manhattan Corp. in New York that offers securities, underwriting, and dealing services in the United States. In 1999, Chase Securities implemented Business Objects SA's BusinessObjects, an integrated query, reporting, and data-analysis tool that works in conjunction with a host of internally developed data mining applications.

"We use BusinessObjects to enhance our daily surveillance and to monitor about 10,000 to 15,000 trades per day," says Ed McLaren, a VP in the compliance department at Chase Securities. "If a problem is detected, we can drill down, see more detail, and then create specific reports on the trades that we deem to be red flags."

Before the implementation, the IT staff at Chase fielded requests for ad-hoc queries from the firm's market analysts, which was an inefficient and time-consuming process, McLaren says. BusinessObjects provides a front-end tool that lets market analysts run static and ad hoc reports themselves--say, on one trader's trades over a day, week, or month. That makes the analyst's job easier and quicker--and it lets Chase's IT staffers deal with other technology issues full time.

"Because compliance regulations are constantly changing, you never know what data you'll need to filter out tomorrow," McLaren says. "Flexibility and ease of use in running ad hoc queries are huge benefits."

Still, data mining and other business-intelligence tools aren't without their glitches. Until recently, critics charged that data mining tools, in particular, didn't integrate well with existing in-house applications, didn't scale to meet the demands of growing companies, and were often limited to one or two modeling techniques or algorithms. Highly trained quantitative experts spent more time trying to access, prepare, and manipulate data from disparate sources and less time actually modeling data and solving business problems using the tools.

But that's finally changing, Smith says. "Products are being built with easy-to-use interfaces, so users can more easily view and understand customer transactional data and detect customer fraud," he says. "Now, any savvy business analyst can use them."

As a result, companies are finding that such tools can give them a tremendous competitive edge, letting them ferret out fraud, cut costs, and better understand their business--risks and all.

return to page 1