Identity Resolution Daily

“A problem well stated is a problem half solved.”
Charles F. Kettering (1876 - 1958)

Have you been wondering how something as big as the sub-prime mortgage fiasco could have not been forecast well before it happened? Ellen Pearlman pondered this while writing in CIOZone about the thoughts of Thomas Redman in his recent book on data quality. Interestingly, Redman wrote that the mortgage crisis “illustrates perfectly how bad or missing data contribute to issues of international importance and the costs and uncertainties that result”, and he wrote it before this past week’s federal intervention.

The data quality predicament in the mortgage arena is one example of a much more ubiquitous problem. Data doubles every 12-18 months, and that includes bad data. Redman estimates that bad data costs as much as 10-20% of revenue. Doesn’t that mean that the earlier in the data life cycle we get a handle on the data quality problem, the lesser the financial impact?

So says Alena Semeshko in a post on ZDNet UK. “I keep wondering how come data quality check still exists as a procedure performed once in a while, rather than as a part of the front-end process? How come most companies start worrying about the quality of your data only when it’s already dirty and in use?”

Identity resolution plays a vital role in data quality applications. Applying identity resolution on the front-end can ensure that error-filled and fraudulent identity information is detected and kept from entering production systems.

How does it work? Take an example of a web application for applicants or new customers. After an online form is completed, the individual’s identity attributes (name, address, etc.) are compared in real-time to a list of known fraudsters to screen out bad guys. The identity info is also similarity searched to discover an already existing master record for that person, and the two can be matched and resolved to one before the data even enters the system.

Redman’s “rule of 10” states that “it costs ten times as much to complete a unit of work when the input data are defective (late, incorrect, missing, etc.) as it does when the input data are perfect.” With bad data just as with software bugs, the earlier they’re found, the more money is saved.

Share This

[Post from Infoglide] California Lottery Fraud

“[N]ow state and provincial authorities are shifting resources to address another rapidly growing problem: lottery fraud committed by retailers. The most highly-publicized case of retailer lottery fraud in recent years occurred in Ontario but cases reported by the California Lottery Commission, perhaps the largest lottery in the U.S., indicate that they have a rapidly growing problem as well.”

Information & Data Quality Conference 2008 Blog: IAIDQ Announces Plans for Certified Information Quality Professional certification

“The IDQ 2008 conference is over (but I’ve a tonne of posts to upload tomorrow sketched out on my laptop). However, one piece of news was announced at this week’s conference which too important to hold until the airport departure lounge….”

Telegraph.co.uk: Standard Life in call for HBOS insider trading probe

“Details of the takeover talks between Lloyds and HBOS were announced by BBC business editor Robert Peston on Wednesday morning at 9am, sending shares soaring from their low of 88p to 215p in an hour. The deal was not formally announced until four hours later. However, in two trades at 8:57am and 8:58pm, two buyers alone acquired more than 20m HBOS shares at 96p. In total, around 160m shares changed hands in the moments before the takeover was announced, making a paper profit of nearly £190m for the traders and leading to suspicion of a leak.”

b-eye.com - Business Intelligence Network - Blog: David Loshin: IDQ 2008 - Initial Observations

“As I am slowly emerging from the cocoon imposed by the needs of getting my new book on Master Data Management out the door, I am able to get back to everyday life, including two of my favorite activities (together!!) - conference attendance and blogging. Anyway, I am currently in San Antonio attending the 2008 Information and Data Quality conference.”

Data Governance and Data Quality Insider: Are There Business Advantages to Poor Data Management?

“However, this got me thinking. Could the inability for the stores to maintain more that 18 months of records actually be a business advantage? How many customers lose the paperwork, or even forget about their road hazard policies and just pay the replacement price? How much additional revenue was this shortcoming actually generating each year?”

Share This