Reproduction of the book by Joshua Bailey Richmond
| "The Richmond Family" |
Resolve Well and Persevere. That's the Richmond motto, and that's exactly what we hope to do on our next project. We have initiated a long-term undertaking, which we feel would be beneficial to many of you researching your Richmond lineages, and we thought we'd tell you a little bit about it. Although there have been several attempts made in the past to make the information from Joshua Bailey Richmond's book "The Richmond Family" generally available to the public on the web, to our knowledge none of these attempts have been very successful. Therefore, we have decided to make a professional reproduction of J.B. Richmond's book (all 600+ pages) in both published and on-line formats. In addition, we also intend to compile all of the data from the book (all 18,000+ names) into a searchable database, which when completed, will be available at this site. Furthermore, this database will be available for download from our site in the GEDCOM format. As you may imagine, this is quite an undertaking, and we expect to be heavily involved in this project in the coming months. But to give you an idea of what the final product will be like, we have included links at the bottom of this page to the first few sections of the book which we have completed.
By way of information, we'd like to share some aspects of the complicated transcription process that is being used to make this project possible. First, each page is scanned using the latest OCR (Optical Character Recognition) software. OCR scanning creates a text file of the scanned image, which can then be edited in any word processor. However, while OCR works quite well on quality documents with standard text, it does not do so well with the pages of Joshua's book. Unfortunately, the old blotchy fonts used in the late 1800's, coupled with the fact that the pages in print today are really just photocopies of an old used book (complete with pen marks, ink smears, check marks, etc.), make the translation full of literally hundreds of errors per page. The italic fonts from the original come out even worse! At one point, it appeared as if it was going be easier to retype the entire document (all 600+ pages) than to fix the errors created from the OCR scanning.
Faced with this dilemma, we decided to write a program in PERL, which works well with manipulating text and data. We wrote into the program a lot of basic spelling and grammar rules, as well as the ability to search for and correct dozens of common words and names. For example, the "Richmond" name kept getting translated as "RicbmonJ" or some other slight variation. The word "born" often comes out as "bom" or "ban." The program automatically fixes all these types of errors. Dates were another problem; the number "4" comes out as a "+" as often as not. Date formats are also frequently incorrect. "Jan. 15, 1844" may come out as "Jan, lS. l8++", with l's for 1's, S's for 5's, etc. The program fixes these types of errors also. It also works to eliminate weird characters like @@$^**(^__-=__+%6 which are usually the result of pen marks or ink blotches on the pages. While the program is not perfect, and in fact in some cases it actually introduces some new errors, it generally does a fairly good job of fixing the majority of the errors (probably about 75%). The program is run on each page separately in an effort to fix as many errors as possible. This process decreases the errors to about 25 to 30 per page (not including the names of spouses in italics since they come out so garbled as to be beyond even human recognition). Additionally, we wrote another PERL program that reads in order the list of page files that belong to each chapter. The program then combines the text of each scanned page into one large file. This new file is then ready for hand editing, which either of us complete by comparing the computer generated document against the original. The editing can take many, many hours per chapter. Once all the corrections are made, the files (which have since been converted from text format to MS Word) are specially formatted, including proper spacing, returns, indents, font size changes, italics, bold, and superscripts. In addition, any images that will appear in the final document are inserted at this time. Finally, the files are translated into PDF format so that they can be viewed on the web.
Phew! As you can see it is a very lengthy and involved process, even with the programs that have been written to expedite the work. The project requires a huge time investment, so please be patient with us as we put this together. To those who desire to assist with the editing, we sincerely appreciate your willingness to help. However, we are cordially declining help with the project at this time as the logistics of file transfers would require additional effort on our part, and distract us from accomplishing the work. We are, however, soliciting help in locating quality images of Josua Bailey Richmond or any of the other images included in his book. If you know of any availble images or photographs that could be used in the new book, please contact us. We are grateful that many of you have found our Richmond Family web page to be helpful, and we hope that this latest project will be of even more value to those seeking out their Richmond ancestry. As a preview of what we have completed thus far, please take a look at the first four generations of the book that we have provided for your viewing as PDF files. Enjoy. . .
 Introduction.pdf | 1st Generation.pdf | 2nd Generation.pdf | 3rd Generation.pdf | 4th Generation.pdf | 5th Generation.pdf |
Sincerely,
Dean Richmond - Genealogist
Todd Richmond - Webmaster
February 28, 2002
You can now dowload the orginal JBR's PDF files at the Richmond DNA Project web page.
Thanks, Jeff.
If you require Adobe Acrobat Reader follow the link.
| If you have questions or data that you would like to share, please email us! | 
|
