Some people eat, sleep and chew gum, I do genealogy and write...

Tuesday, February 10, 2015

BYU Family History Technology Workshop -- Developer sessions

Gordon Clarke of FamilySearch presenting
In the afternoon session we had the following:

Source Linker: Bridging the Evidence by Randy Wilson from FamilySearch

My reporting and impressions:

We used to have GEDCOM for exchanging data between programs. Genetech Geneealogical Data Model, (GDM) has persona or information about a person from a single source. New paradigm is to model what the sources say and represent that. Persona linking advantages include additional data and relatives, shows conflicting data, engenders conclusions based on all the data. We need to pull the data out of the images. Use records to build a tree. (This is exactly what I am in favor of and teach). Records are used to build the tree. The complication comes with multiple personas as we add sources with additional information about the sources.

Source Linker is a program that links the persona to a tree person. So, the persona is what is extracted from the sources that becomes the linked person in the family tree. This is what is happening in Record Hints in Family Tree. Source comes into the Family Tree with a citation. The process is to review and attach those records that are evaluated as correct. This is the process that is being used by the Family Tree to match sources. Source Linker is giving a basis for adding individuals from records which I heartedly support.

Evolving Family Structures, Representation and Visualization by J. R. Hott, Computer Science, University of Virginia.

My reporting and impressions:

Chord and Flow Diagrams from the Nauvoo Community Project by Jill Crandell. Using individuals selected for the project. Researchers need to ask a series of questions about the relationships. In his case, Brigham Young's marriages. What is a marriage etc?

Chord diagram of Brigham Young. This is an interesting way of representing a rather complex system, i.e. the family of Brigham Young. To see how these look, click here. To see more about this presentation see Geneology Visualizations. You will find more information also at this link.

New Apps, new Gallery, New Future by Gordon Clarke of FamilySearch

My reporting and impressions:

New App Gallery from FamilySearch. Despite trouble with sound on the computer, (never happens to me :-), this is a very interesting subject. FamilySearch does need progress and innovation. Mobile has out grown desktops, which I have pointed out several times in the past. Social media has surpassed email. Email surpassed snail mail a long time ago in 2009. Mobile computing is mostly games, social networking and entertainment. FamilySearch needs more apps, fun and media.

The idea is to make family history more engaging for more people. More records used more people, i.e. ancestors, discovered. This is done through FamilySearch APIs and other connections. The new future is social, mobile, search-ability, DNA, medical and finding help.

The Future of Family History Technology: Challenging Opportunities and Data Sets by Jake Gehring of FamilySearch

My reporting and impressions:

All FamilySearch imaging is digital presently and now records are ready when you are. We need better coverage around the world. It used to be that genealogy was only for the rich and prominent. Now we have a lot of compiled genealogies for average people. Now we have research conclusions in digital format online in family trees and again the content is ready when you are. In the long view there have been tremendous advances in genealogy.

Computers are tools to understand information not just represent that information. We don't just express ideas but also manipulate and interpret ideas better with computers. FamilySearch acquisition digitized about 130 million images in 2014. We can improve digital images.

It was interesting that Jake mentioned an experience in doing the research about a family, putting the conclusions in and then having a family member change the information back to inaccurate information. Yes, we have a long ways to go. FamilySearch only indexes about a quarter of all the records acquired every year. Unindexed records are not searchable. The issue is the rate of Indexing.

FamilySearch is working on natural language processing to extract information from records. They are looking for systems that can correct errors. We need to make it easier to interview your relatives.

No comments:

Post a Comment