Skip navigation

Published May 30, 2021

| 4 Comments | Leave A Reply


Last week’s “Roots & Branches” column took a look at the efforts of FamilySearch.org, the free genealogy website owned by the Church of Jesus Christ of Latter-day Saints, to avoid a “dark archives” in which digitized images of records are simply plopped up on the Internet without further description.

FamilySearch already has some 4.5 billion images online and with more than a million images added daily—it’s a lot to keep up with!

But even stemming that tide by keeping up with processing and describing new materials going live on the site is only half the battle as far as genealogists are concerned.

In an event for genealogy community “influencers” earlier this month, FamilySearch’s John Alexander, a senior product manager, reviewed and previewed what’s going on with making such records searchable.

He noted that extraction of names from records began manually in 1948 and computerization led to the first big project, an index to the 1881 British Census, being produced on CD-ROM in 1998.

A decade later in 2007, FamilySearch web indexing was launched (with an updated format in 2018) and included some 350,000 volunteers who’ve collectively indexed 500 million images.

In 2012, FamilySearch was part of the 1940 U.S. Census Community Project in which some 343,000 volunteers from societies and organizations indexed 797 million records in just three months. FamilySearch made a special push to help get the job done.

With the “hunkering down” caused by the COVID-19 Pandemic, FamilySearch’s indexers—who had primarily been from North America—became a more international force, with 43 percent coming from around the world. And they indexed more than 400 million records!

Alexander said, of FamilySearch’s 4.5 billion images online, only 1.4 billion are indexed. “It will take 170 years to index all our current images at that rate,” he said, “and due to faster image gathering, the backlog grows by 20 years annually.”

Currently it takes about two years from the beginning of an indexing project until it is online, according to Alexander. “Our business goal is to get in a record, have it transcribed and indexed and online in seven days.”

How to accomplish such a daunting goal? Enter computer indexing with artificial intelligence.

FamilySearch has computer programs that recognize and pick out specific types of information (i.e. names, gender, dates, places, and other useful information) and provides more useful details than human indexing provides, Alexander said.

 “As computers get better, they will index records in hours that would take centuries for human indexing,” Alexander said. “Records already uploaded are marked by a blue banner. The images are available and edits are possible. To a degree, errors patrons can find are fixable.” He said both the original and “corrected” information is kept for future patron reference.

All in all, it’s obvious that FamilySearch has an overwhelming commitment to making records available and searchable to genealogists—and that they have spent the Pandemic over the last year innovating and moving ahead, a silver lining for what has been a fraught time.

4 Comments

  1. Donna Jones

    4 years ago  

    The technology and commitment of FamilySearch is amazing. Thanks for the interesting post.


  2. Eric M. Bender

    4 years ago  

    I used to spend hours — days, weeks! (and a small fortune)— in Pennsylvania going through files, books; scrolling miles of film. I’m still compiling the information. Why? It’s virtually all available now online.

    Was I born too soon or too late?! — Rick