I hope someone who works for FamilySearch reads this because I believe that I’ve got a great idea. The idea I have in mind is to automate the indexing process using computer programs and Artificial Intelligence. We would be able to index so many more documents in a much shorter time period if we could develop a program or device that could scan documents and automatically/instantly index them. This increase in indexing efficiency would allow us to index so many more documents, and therefore we would also need to find more documents to index in this faster manner. After all, this is the Gathering of Israel we’re talking about, and we should be trying to do this important work as fast as possible!
Not sure why this is a duplicated question in Indexing, but, I will copy and paste the comment I made on the other post:
BYU's Computer Science page on AI and Machine Learning is a fabulous source to see all the great accomplishments that these students and faculty are making in the area.
A few years ago I participated in a project where we were highlighting stories from newspapers just to teach the computers how to read deaths, marriages, and births so the computers could learn how to index them.
Now there is a project called Reverse Indexing. I don't know how long they plan to keep it up, but, it is fun to work on!
Here is a quote from an earlier post from the BYU Linking Lab with a Reverse Indexing Update:
"If you haven't tried this application yet take a few minutes to watch the video and do a page or two. it is easy and it helps the computer learning for future indexing projects."
"Many of our ancestors were occupational pioneers. Wouldn't it be great to help people see if there ancestor was the Black police officer or the first female lawyer in a particular town? To make this possible, we are indexing the occupation, industry, and education field in the 1940 census. Our machine learning algorithm is doing really well most of time but needs some help from human insight. We've created a short video that shows how you can help. We are making a huge push on this project over the next few weeks so that we can deliver the data for these fields to FamilySearch."
Here's a video about this project: https://youtu.be/DZUnNHyhkDk
I think you will enjoy visiting these sites, @IsaacHemingway !0
This would be something you might want to repost into the Ideas section of Community. You can find that by the button on the left side of your screen.0
@AGTHBG It is really weird because the duplicate post has a voting box like it is in Ideas, but, it isn't - it is in Indexing. Really strange.0
Interesting. Not sure why it did that. There are a lot of changes happening in Communities so there is going to be clean up work in the software. Perhaps that is why?2
There are some old books that have been scanned online then AI is used to transcribe to make Google searchable. It does okay for typewritten things, but even when something is typed, it still has lots of errors. Not sure if it could do a lot of the poor condition documents with varying types of handwriting. Maybe this could be something used in tandem with a human with critical thinking ability, but it seems there are a lot more people interested in indexing for the first time than arbitrating and going over everything again to make sure it is correct.1