Posted on 04/21/2004 11:27:19 PM PDT by neverdem
LARS LUNDIN and his database are matchmakers of a sort, trading in information about matches made long ago. Genealogists send him family trees, and his computer compares them with other family trees in a database of more than two million current and former Danish citizens. Each month, he estimates, his computer finds more than 1,000 connections between cousins.
Projects like Mr. Lundin's are becoming increasingly common as both amateur and professional genealogists use the Web to pool resources, compare notes and connect family histories. Some, like Mr. Lundin, do it for fun, while others are building companies that offer higher-quality resources to paying customers. All of them are taking advantage of the plummeting cost of storing information about billions of people, living and dead, as well as the Internet's ability to lower the barriers to cooperation.
Mr. Lundin's work began as an experiment when he was a graduate student in applied mathematics at the Technical University of Denmark. A fellow amateur genealogist suggested that it would be impossible to create a database that contained information on everyone who has ever lived. This attitude led him to start calculating.
"The total number of people to have lived in Denmark in the genealogically relevant period from 1600 to 1900" did not exceed 10 million people, he said, and he concluded that this was "an entirely manageable size."
Now those who want to research their Danish heritage can send Mr. Lundin a family tree in a format called Gedcom (from genealogical data communications), and his software will search a collection of family trees from other contributors for shared ancestors and cousins. The Gedcom file format, originally developed by the Mormon Church, is the lingua franca of the genealogy community; it allows users to convey information about ancestors, their birth and death dates and other information. Mr. Lundin's software looks for the best matches with other Gedcom files. When it discovers that two people have ancestors in common, Mr. Lundin said, he writes to both of them.
Mr. Lundin works in a corner of an online world in which information about billions of people is being traded through free and subscription Web sites, bulletin boards and data clearinghouses. One of the most popular free Web sites, www.rootsweb.com, is owned and managed by MyFamily.com, which also runs one of the largest pay sites, ancestry.com.
People can use Gedcom files at Roots- Web. com to find common ancestors in much the way that Mr. Lundin's computer matching service does; RootsWeb also runs many bulletin boards and mailing lists for people who are interested in particular family names, cities or regions.
Craig Sherman, senior vice president of MyFamily.com, said the company supported the free information on RootsWeb, and other projects like FreeBMD (www .freebmd.org.uk), a volunteer registry of British births, deaths and marriages, to encourage enthusiasm for the hobby.
"RootsWeb is a great introduction to Ancestry.com." Mr. Sherman said. "It's one of our most cost-effective sources for new subscribers to Ancestry.com. It's a great retention tool because when people are doing their family research, one of the best ways to be more successful is to go to the Roots- Web message board to find people who are researching similar family lines."
More than two billion records, from thousands of sources like the United States Census (1790-1930), immigration records and Civil War pension records, are cataloged at Ancestry.com, he said. The company uses the revenue from annual fees ($40 to $200) paid by its 1.6 million customers to digitize new sources. In March, for instance, hundreds of new databases with names like "Pike County, Missouri, Deaths, 1878-1904" and "Thornhill, Yorkshire, Parish Register, 1580-1745," were added or updated.
Pairing volunteers with professionals lets each take on the work for which they are best suited. The volunteers handle questions that require focus, shared interests and creativity, while the fee-based site specializes in brute force requiring thousands if not millions of hours of work.
The volunteers who contribute to Roots- Web and other sites reap their own rewards. Gayle Triller, a supervisor at a cellphone company, lives in Salem, Ore., and runs the RootsWeb bulletin boards for Baltimore, the state of Maryland and several counties in Texas and Mississippi.
Ms. Triller said she began the pages, which are filled with phone numbers and addresses of local resources, because she had ancestors who lived in those places. "My mother's family is from Baltimore," she said. "My dad's family is from Mississippi and Texas." Running the sites enables her to trade favors with residents who can visit local offices for her.
Such legwork is important because even though database searches are a good way to find people with the same name, the results are often spotty and incomplete. For many genealogists, the initial gratification of uncovering large portions of a family tree is followed by the realization that links must be double-checked and the mistakes of others must be corrected.
The amount of work involved in verification depends upon the user's standard of proof. "I have never discovered provable family through digital genealogies," said Steve Hall, an amateur genealogist in Apex, N.C. "But I have found leads."
Mr. Hall said he relies on primary sources like paper documents, because he has found that errors creep into electronic sources. The demand for primary sources explains why some sites like Ancestry.com now post reproductions of the original documents.
Like many genealogists, Mr. Hall uses a personal database to store the information about his relatives that he gathers. More than a dozen different genealogical database programs are available from a variety of vendors. Mr. Hall uses the Genealogical Research and Analysis Management Programming System, or Gramps, a free program developed by Don Allingham. Like other such software, Gramps (gramps .sourceforge.net) enables people to enter data about their ancestors and store the results in a Gedcom file; it can also produce charts, documents and Web pages from the information.
Gramps, like many other versions of genealogy software, is a collection of programming contributions from volunteers. A part written by one contributor, for example, produces pretty reports, while other parts, written by other people, can comb through accumulated data and perform what Mr. Allingham calls a "sanity check" by flagging factors like ages that do not match, marriages or births involving people too young or too old, and other obvious glitches.
Such tools make it easy to put together extensive family trees. Andre Brummer, a vice president for products at MyFamily .com, inherited genealogical research collected by three previous generations and assembled it into a computerized tree containing more than 22,000 people, with some lines stretching back to the 1600's. He has traced his family line back through South Africa, where he was born, to Germany, France, the Netherlands and Britain.
"Most of our work was done the hard way," he said. "When I started, we did everything by hand. Trying to coordinate all of that and keep it together was a mammoth task."
That task is much easier now.
Thadd Vargas, a pharmaceutical licensing representative, decided to experiment with Ancestry.com's World Tree by uploading a simple Gedcom that went back six generations and found a link with a deeper tree.
"Some of my information was incorrect, but the software suggested the link anyway," he said. "When I looked at the details, it was a match, and the tree I was linked to took me back to Salem, Mass., in the early 1600's. A little more work like this and I was back to Castle Leicester, England, in 1455. All in a day."
I have no idea, but thanks for reminding me to send this story to my cousin.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.