Data Integrity

views updated May 14 2018

DATA INTEGRITY

Information is what gives power to the Web and e-commerce. In a matter of minutes, consumers are able to research, compare, and purchase products and services online. This availability of information has created more discriminating consumers and has put increased pressure on retailers to offer competitive prices. Understandably, the accuracy or integrity of data is very important during e-commerce.

SOURCES OF FAULTY DATA

The integrity of data can be compromised in a variety of ways, including malicious proprietors, human mistakes, and technical error. Unfortunately, like accurate information, faulty, inaccurate, or misleading information also travels freely on the Internet and consumers and companies alike have been negatively affected by it.

Fraudulent business schemes account for a significant amount of erroneous information on the Web. In this scenario, sellers often make faulty or exaggerated claims about products and services that sometimes do not exist. According to Nua Internet Surveys, a study by the Worldwide E-Commerce Fraud Prevention Network found that 50 percent of businesses in the United States saw online fraud as a significant problem. Ten percent of those companies ranked online fraud as their "most significant problem." Among those surveyed, half had experienced losses between $1,000 and $10,000, and 19 percent in excess of $100,000. Online auctions were among the leading areas of e-commerce fraud in the early 2000s. Nua Internet Surveys reported that the U.S. Internet Fraud Complaint Center (IFCC) registered more than 20,000 complaints in its first six months of existence. Sixty-four percent of the complaints involved auction fraud.

In addition to information that is simply unreliable or inaccurate, the integrity of data can be compromised in other ways, including technical errors that happen during data transmission. Along with the growth of e-commerce came an increasing reliance on software programs that automate tasks involving databases of customer information. This opens the door for computers to accidentally execute tasks that affect thousands or tens of thousands of people at a time. Before the advent of e-commerce, mistakes usually involved a handful of customers, or at least could be caught in time to avert a major disaster. Although consulting firms, software developers, and leading companies all devote resources to this problem in their own ways, technical error was still a major concern in the early 2000s and no one solution existed that ensured the integrity of information during all Internet transactions.

A more simple source of technical errors, which some e-commerce developers were dealing with in a variety of ways, involved limitations of Web technology. Traditionally, Web servers (the computers or software responsible for maintaining and storing Web sites) break off connections with clients (individuals using Web browsers like Microsoft's Internet Explorer) after a Web page has been downloaded to their screen. This means that when consumers view a Web page, it may become outdated in a matter of seconds. Products that were available when the consumer initially downloaded the Web page may become unavailable several seconds later.

Errors were also attributed to online shopping carts—the technology that keeps track of items consumers are interested in until they are done shopping on a Web site. For example, if someone removes an item from their shopping cart before checkout because they don't want it and then clicks on their browser's "back&rquo; button to revisit a Web page they just saw, it is possible that the unwanted item will be added back to their cart. Other similar sorts of mistakes can happen with online shopping carts.

Human mistakes as well are sources of compromised data. In October 2000, Buy.com agreed to a $575,000 class-action settlement when, due to a human data-entry error, it mistakenly priced a computer monitor at $164.50 instead of $564.50. Buy.com honored the price for the monitors it had in stock. However, when it refused to fulfill all of the orders it received, consumers sued. Around the same time, consumers threatened to sue Egghead.com for a similar blunder. The retailer mistakenly priced a $335 computer component at $34.85. Before the mistake was noticed, Egghead.com had received thousands of orders, which it cancelled.

CONSEQUENCES OF FAULTY DATA

The consequences of bad data are real. Chain Store Age Executive with Shopping Center Age, revealed how a lack of data integrity can cost companies money, explaining: "At the most fundamental level data error attacks those things that affect revenue, costs, and ultimately, customer loyalty. Unanticipated out-of-stocks equal lost sales. Excess safety stock requires higher investment and more mark-downs. Inaccurate inventory data demands that resources be allocated to determine the availability or shortages of items. Data error also feeds invalid information to the inventory replenishment system, driving poor merchandise planning and purchasing decisions, resulting in more obsolete and excess inventory."

The Chicago Tribune listed several serious data errors that serve as examples of the kinds of things that can happen when large volumes of information is digitized, stored, and transferred. Among them were a credit reporting company that accidentally labeled 1,400 residents of an Eastern town as having bad credit and a health insurance company that accidentally sent out $60 million worth of duplicate checks after its new computer system malfunctioned. Finally, Legal Assistant Today described how an engineer at PairGain Technologies Inc. created a false Web page announcing that it was about to be acquired by an Israeli telecommunications firm. The page's design was similar to that of a popular news service, and it falsely motivated investors, causing the company's stock to increase by 31 percent.

ENSURING DATA INTEGRITY

Although there is no universal way to ensure the integrity of data and technical errors are bound to occur from time to time, consumers and companies alike are able to take measures to protect themselves. Common sense and critical evaluation were essential first lines of defense for both parties. For consumers, measures could be taken to verify the identity of a Web site's owner. Secure certificates issued by companies like VeriSign and CyberTrust were means of doing this in the early 2000s. Additionally, consumers were able to limit their transactions to those companies providing encryption methods, whereby information was scrambled between the sender and receiver.

Companies engaging in e-commerce had several methods of ensuring the integrity of data. Simple measures involved limiting the number of parties responsible for posting data to their Web sites and building in extra layers for fact checking. Special software programs for detecting and correcting errors were likewise available.

Companies also developed elaborate data management strategies that involved reviewing the ways data was collected, stored, updated, and used. General Motors was completing a customer data quality project in late 1999 that involved "capturing customer information from multiple sources in a standard, error-proof way, then merging it with detailed demographic and lifestyle information," according to Information-week. Over time, some companies find themselves with a large number of different databases across the organization. A data management strategy might involve consolidating this information in one secure place and identifying the best ways to integrate it with different processes and systems.

data integrity

views updated May 29 2018

data integrity Resistance to alteration by system errors of data stored in a computer. It is a condition that denotes only authorized and proper alteration of data. It is a measure of the reliability of data read from magnetic media, in terms of the absence of undetected errors (see error rate). However the undetected error rate perceived by the host system may be worse than that arising at the magnetic disk or tape if undetected errors can arise, e.g. from the effect of noise on connecting cables where the interface concerned has insufficient error detection capability.

From a system point of view the undetected error rate of a peripheral may be inadequate: the system can improve on it by making additional provision for checking in software.

A Dictionary of Computing JOHN DAINTITH