Posted on: Sunday, April 11, 2004
Firms fighting plagiarism with software
By May Wong
Associated Press
White-collar copycats may be less inclined to pilfer the well-chosen words of others now that software designed to ferret out plagiarism is moving out of academia and into the business world.
For years, educators at colleges and universities have marshaled software tools to ensure that their students' work is original.
Now, tainted by scandals or leery of the Internet's copy-enabling power, a growing number of newspapers, law firms and other businesses are using data-sifting tools that can cross-check billions of digital documents and swiftly recognize patterns in just seconds.
Unlike Google and other search engines that find matches to typed-in key words, an advanced plagiarism-detection service such as iParadigms LLC's makes a digital fingerprint of an entire document and compares it against material on the Internet and in other sources, including proprietary academic and media databases.
Even the U.N. Security Council has begun to protect its credibility this way, using iParadigm's technology since last fall to ensure the originality of reports by its researchers and freelance writers.
Silent searches
Oakland, Calif.-based iParadigms started in 1996 with a computer program to help researchers at the University of California-Berkeley inspect undergraduates' papers. Today, its Turnitin plagiarism-detector is used by about 2,500 high schools and colleges in the United States and 1,000 more abroad. It launched a commercial version, iThenticate, in January.
Other plagiarism-detection providers, including Glatt Plagiarism Services, MyDropBox LLC, and CFL Software Development also report growing business outside the educational sector.
New clients include companies that produce instruction or training materials, attorneys searching for copyright violations, Web sites and police and military agencies that check officers' applications for promotions.
Few of these businesses are willing to talk about using these tools. Many insist that the software makers shield their identities and keep mum about any transgressions that are exposed.
None of the plagiarism-catchers said their clients were willing to be interviewed for this story. Fearing negative publicity, most "don't want other people to know they're using the service," said Max Litvin, co-owner and inventor of MyDropBox.
Last year, one publisher turned to iParadigms when it investigated — and subsequently affirmed — rumors that an accomplished textbook author had plagiarized other sources. Sworn to secrecy, iParadigms president John Barrie said he watched in disbelief as the publisher quietly revised later editions, leaving the author's reputation intact.
"But I see a lot of plagiarism every day," Barrie said. "Most authors, whether a student or professional author, they think the odds of being found out are so remote that they'll play the odds and think they're just fine."
Cut-and-paste tactics
IParadigms charges universities a $500 annual licensing fee plus 60 cents per full-time student. Business customers pay $1,000 a year and $10 for each page submitted for screening. Newspapers face different charging options based on word count or circulation.
A different program, WCopyfinder, was employed by USA Today as it probed the work of its former reporter Jack Kelley. The free program compares strings of words only from preselected documents.
IThenticate and MyDropBox, by contrast, are Web-based tools. Users upload documents to the Web sites; the services troll the Internet and other proprietary databases, such as Lexis-Nexis or ProQuest, for any sign of unoriginal work; then they produce reports showing matches. IThenticate also combs its archive of Internet pages, which grows by 40 million pages a day.
Clearly, plagiarism is a growing problem. In a survey of 30,000 undergraduates at 34 colleges, 37 percent admitted committing cut-and-paste plagiarism using the Internet, up from 10 percent in 1999. Only 20 percent of their professors use plagiarism-detection tools, according to the survey by Rutgers University professor Don McCabe, founder of the Center for Academic Integrity.
Plagiarism detectors can be relatively cheap insurance against intellectual property sins, but many businesses and even educators remain reluctant to use them. Some fear lawsuits if they accuse someone of cheating. And deciding what amounts to actual plagiarism remains a judgment call that humans must make, creators of the software say.
"It's merely a tool to guide the eye," said Lou Bloomfield, a University of Virginia physics professor who created WCopyfinder in 2001 to check for plagiarism in student term papers.
Loss of trust
IParadigms software helped the Hartford (Conn.) Courant conclude last month that Central Connecticut State University's president, Richard Judd, had committed plagiarism in an op-ed piece after an alert reader said it may have lifted sentences from the New York Times.
The newspaper tried an Internet keyword search but without much success. IParadigm's software later showed that the opinion piece included not only material from the Times but also three other sources — at least 11 percent of it appeared to be unoriginal.
The criticism upended the respected university administrator's career — Judd, 66, announced on March 19 that he will retire July 1.
The Courant doesn't plan to routinely check every story for plagiarism — just submissions for the editorial page, says John Zakarian, editorial page editor. However, the paper now has a fast and effective tool to use if a staff writer's story is questioned, he says.
"We've come to rely more and more on the Internet," he said, "and it's not humanly possible to verify every sentence and word. I was amazed we have the wonders of technology to help in that fashion."
Other newspapers are reluctant to use the powerful software.
At the Macon (Ga.) Telegraph, which fired a reporter for plagiarism last month, editors are discussing how to prevent a repeat occurrence — but such electronic tools aren't being considered, said managing editor Mike McQueen.
"We, the editors, trust our writers deeply," he said. "I don't think anybody here would want to challenge our reporters to prove that they are not plagiarizing everything they write. It'll look like a witch hunt."
Still, Barrie predicts that iParadigms' commercial clients will eventually outnumber the academics. "The stakes are 100 times greater," he said. "We're not talking about grades anymore."
