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

Wednesday, February 11, 2015

Innovator Summit - Mission Impossible? Tough Challenges, Which Need Automated Solutions

Eastside View of the Salt Palace, Salt Lake City, Utah
We had a nice box lunch and I went to the Family History Library for a short break to work with some of the FamilySearch Research Wiki participants. Nice to see familiar faces and talk about the challenges facing the Research Wiki. Back to the Innovator Summit and a forum with four presenters:
  • Robert Charles Anderson, FASG 
  • Diane C. Loosle, Family History Library and Centers Director, FamilySearch 
  • David E. Rencher, Chief Genealogical Officer, FamilySearch 
  • John Wylie, Member, GenTech Genealogical Data Model Working Group
The title of the presentation is in the title to this post. This is the link to the syllabus. As usual, I will be reporting and making comments.
Panel Presentation. Yes, the rooms are this dark. 
Diane Loosle:

Toughest genealogical research challenges we face today in genealogy. We have places to put conclusions but not a lot of programs for putting in the rest of the problems. Example, "find the father of John Smith." In the survey phase, we are trying to determine what has been done previously. Identify and select sources. Need to put the information in a collection and provide links between data. Share the repository.

There is nothing in programs about patterns, i.e. birth order, naming, migration etc.

We need a workspace for all of the and place to gather all the information about the same person or persona.

Comment: I certainly agree that these area are not currently covered by the existing programs.

John Wylie:

He speaks of the following:
Step 1: Lexicon (not yet available)
Step 2: Data Model - GEDCOM X, Gentech
Step 3: Tools - Mostly tree systems available, many more needed

We will not compromise the genealogical process to accommodate technology. Modeled as they understood it. Did not know the future. Technology has moved beyond our imagination. The data model is in the public domain.

Robert Anderson:

Explains some of the issues in creating the Great Migration Study Project. "What do I have to believe to disbelieve the conclusions." Comment: argumentum ad absurdum. He discusses linkage analysis. See the topic of genetic linkage. Also see Robert's book below.  Linkage data between sources and conclusions are missing from the present programs.

David Rencher:

Creates a spreadsheet of the data in a parish register or census. Sorts the list by given name and other topics such as witnesses. This gives a way to analyze the data. He is looking at name patterns and birth order and other data sets. They are asking for a way to organize additional outside sources that are not presently available. We never ever start with the tree. Sources should not go straight to the tree. Need to pull out all the data points that pertain to a certain person including land records, neighbors, friends, relatives of the target person. Need to build tools to fill in the missing steps of the research process.

Diane Loosle:

Use of timelines to differentiate individuals from the same area by examining a case study. Her whole project took about sixteen years. This is the same way of using a spreadsheet to look at and verify individual entries. These are steps for which we do not presently have program support. Need a data collection repository separate from family trees. Posit criteria for evaluation. This is an excellent example of using a set methodology to distinguish individuals in the same area with same or similar names. You have to look at a lot of individuals who are not going to land in my family tree. Needs a great deal of research and data. We need to know when data is added which matches our research.

We also need to have ways of correlating data that is inconsistent with historical facts. Need visualization on maps.

David Rencher:

650 of the 3500 counties in the United States have some level of record destruction. We need to have a visualization tool that shows us when records are  created. Comment: This is a good argument for a source-based research system.

Diane Loosle

Challenge is to grab a huge amount of data and establish data sets that gather people who may be friends or relatives in the same area. We need certain tools that are presently not available.

  • Metadata on collections showing what is not included
  • Tips to the user to warn them of reasons why their person may not appear
  • Suggestions of other collections

David Rencher:

We need a way to unravel mistakes made in the trees. The ability to undo what was done to see when mistake was made.

Now there was a question and answer session: We are begging people to innovate.

Reading list from the syllabus:

  • Anderson, Robert Charles. Elements of Genealogical Analysis. Boston: New England Historic Genealogical Society, 2014. 
  • GEDCOMX. “GEDCOMX.” Accessed 1 December 2014. http://www.gedcomx.org/. 
  • Jones, Thomas H. Mastering Genealogical Proof. Arlington, Va.: National Genealogical Society, 2013. 
  • Leary, Helen. “Evaluating Research Data” and “Designing Research Strategies” in North Carolina Research: Genealogy and Local History, 2nd Ed. 3-67. Raleigh, N.C.: North Carolina Genealogical Society, 1996. 
  • Morgan, George G., and Drew Smith. Advanced Genealogy Research Techniques. New York: McGraw-Hill Education, [2014]. 
  • National Genealogical Society. “References for Researching: Genealogical Data Model: The GenTech Genealogical Data Model.” Accessed 1 December 2014. http://www.ngsgenealogy.org/cs/GenTech_Projects. 
  • Osborn, Helen. Genealogy: Essential Research Methods. London: Robert Hale, 2012.
Looking down at the Innovator Summit area at midday

No comments:

Post a Comment