Kentucky Record Transcriptions: A Case Study in Computer Assisted Profile Creation

I will be describing everything in EXCRUCIATING detail, but for now a quick summary, with no attempt to make it digestible for the non-specialist:

The book Afro-American deaths of Boone thru Boyle County, Kentucky, 1852-1862 by Gwendolyn Tippe is basically pages of data in a tabular format. I downloaded pages as JPEG images from FamilySearch, then used the OCR program tesseract to produce text files from them. I hand edited the resulting files so each row was a comma separated value list. I processed the edited text files using a program I wrote in Python to produce two new files. One a .csv file I can import into the GRAMPS genealogy program. I then exported the tree I created in GRAMPS to a .ged file, which I import to WikiTree. The second file I fed to a program I wrote in Python that generates the full text for the bio section. As I added the profiles through the GEDCOMpare process, when I got to the step of editing my new profile all I had to do was copy and paste my generated bio into the text box. BOOM! Done.