Crowds gathering for the first day of RootsTech 2019 |
First of all, I attended the Access and Preservation sessions this morning. These sessions were developed to help archivists learn more about what various archives are doing to make their records more available and useful to the public. David Carmichael, archivist for Pennsylvania, commented that archivists know what they do but many don't know why they do it. They need to talk more with the users to see how the materials they have in their archives are being used and hear the stories those records provide, not just know that they have so many linear feet of records on a particular subject. Mr. Carmichael also discussed the different problems they experience with storing digital records versus the physical records. Many archives have records that are several hundred years old but often have difficulty maintaining digital records in a usable format over the period of a few decades. Back in the 1980's his archive scanned many of their collections in a digital format developed by IBM but that format is no longer supported. Luckily they had maintained their physical collections and could scan them in a more modern format. Now they store their records in PDF-A format which is an archival pdf format and should last for the foreseeable future. Another problem they have today is that many modern records have never had a physical presence. These records, such as e-mails and digital photography, have only been digital. Several positive features of digital archives is that they can be monetized and there are companies that are producing digital records for the marketplace at a rapid pace. Additionally, digital records allow cooperative efforts and virtual communities to develop. Groups of people can get involved by providing and indexing collections that might not have been available otherwise.
Joe Price of the BYU Linking Lab discussed several projects his students are working on. One project is working with computer AI to learn handwriting recognition. The billions of digital records that are currently available can be compared with what humans have indexed so the computers can begin to learn to recognize the handwriting patterns. Another project is using these large record databases to begin statistical analysis of communities or groups of people for scientific studies. This moves the records that we think of as genealogy records into a new realm of users which may provide us more insight into their lives. For example, they are taking the 1850-1940 census information and the computers are linking them together into a census tree to follow people and their migration patterns. This analysis can follow individual families across the range of census records and also link them to other records such as school yearbooks and directories to gather more information. They are now linking records from the major genealogy databases to build the stories of individuals. This helps to find people who are not yet in FamilySearch's Family Tree through Tree Extending Hints. One example of this is a webpage called the Social Networks and Archival Context (http://snaccooperative.org).
The third talk in the Archives and Preservation series was from Kris Rcepzynski, archivist for Michigan. He talked about the seekingmichigan.org website and how their partnership with FamilySearch has provided increased access to various Michigan records. This website is undergoing a major facelift and will re-emerge later this year as michiganology.org. They are currently working on putting up county probate records as well as county naturalization records. They currently have naturalization records for 70 out of the 83 Michigan counties online.
During lunch I attended the FamilySearch talk. They discussed the three major problems they are experiencing.
1) My country's records are not a priority. - Priorities are being set by the number of requests they have gotten for various records. Some areas have very few requests and are therefore lower on the priority list.
2) There is a backlog for indexing records. - The current backlog is 40 years. By the end of 2019 the backlog will be 50 years and by the end of 2020 it will be 60 years. This is due to the large number of new records that are being digitized. We either need to find a way to increase the number of indexers from the 300,000 we currently have or find a new way to index records.
3) We aren't getting records fast enough to avoid loss. - Every year more records are lost due to various reasons. There just aren't enough crews out there indexing these at risk documents to avoid losing some.
So, what is FamilySearch doing to address these problems? Imagine is everyone, from individuals to public archives, had the ability to upload any documents they have and those records could be processed, indexed and published all within 24 hours. That would provide unlimited record access and allow communities to work together to preserve their own sets of records. This is happening in several test areas now. Another advancement is the ability of computers to perform Optical Character Recognition (OCR) of not just typed documents but also handwritten ones. Last year in a test of the computer's abilities they were able to OCR 50 million obituaries in only 3 weeks. This is now the GenealogyBank Obituary Collection we find on FamilySearch. They were able to use the Amazon servers to perform this task in record time. Now FamilySearch users are able to go in and fix indexing errors from this test. They are also working on a new image search process where the computer will estimate where a record should be in an unindexed collection of digital images based on the information you put in the search process. This will save countless hours of searching image after image trying to find your ancestor. Once you find the proper record you can attach it to the person as a source and index it right there.
Martin Luther King III speaking live at RootsTech |
Devon Noel Lee interviewing me at RootsTech |
What else happened? Well, I was interviewed by Devon Noel Lee for her Family History Fanatics YouTube channel. We discussed doing research in the European archives. I also manned a table with several of my fellow members of the Second Life Virtual Genealogical Society. We were talking about the benefits of becoming a member of SLVGS to people who came by. All in all, it was a very productive day and I learned some new things about the genealogy world. But now it is time to get ready for some sleep so I can do this all again tomorrow. Good night all.
SLVGS crew manning a table at RootsTech |