Not Me If I Have Anything To Say About It
As the time-honored axiom goes, a job not worth doing is not worth doing well.
Imagine you have two tall filing cabinets full of minutely detailed roadway records dating back to 1926. Periodically, the state legislature asks you for snippets of data from some of these records, which means you have to dig through mountains of paper to find the specific snippet that's been requested. Gradually it begins to dawn on you that you might like to have this data in electronic form.
But these are pages and pages of columns of numbers in tiny type. The earliest records, of course, are hand-written, then typewritten. As you get on up into the 80's and 90's the papers are clearly computer printouts, which is pretty frustrating when you consider the implication that this data did once exist electronically. Some government employee - presumably someone working for the same department that's heading up the digitization project now - apparently thought it was a better idea to make a whole bunch of printouts, put them in a file and throw the tapes (remember computer tapes?) away.
We have some optical scanning software which is supposed to read all the columns and columns of numbers, preferably with minimal human help, so they can be converted to Excel and eventually stored in an Access database. But it can't read the handwritten stuff, of course. The typewritten stuff is uneven and blotchy, and the computer printouts were Xeroxed down to 8.5x11 size from big green-and-white computer paper (remember that stuff?) I have a brain (or so I often like to tell myself) and I can't puzzle out a lot of the numbers. The optical recognition software stalls at pretty much every single digit, so yesterday I spent my entire day 10-keying in a 6-page report.
My boss was not pleased. She figures, I guess reasonably enough, that she doesn't pay us the big bucks to be 10-key operators. (She pays us the big bucks to go on extended breaks and to do the crossword and sudoku puzzles, but that's a different story.) So she called a meeting with the head of the department in charge of the digitizing project - who has borrowed one other guy from my department and me to help - and he was eager to reassure her that it would take hardly any of our time; all we need to do is look over the documents, set aside the ones the optical recognition software can't read, and skip over the errors caught in the documents that it can. All that stuff will be taken care of in Phase II, the quality control.
Um. In other words, the whole project. Who do you think they're going to tap to do "quality control"?
posted by Beth @ 6:24 PM
3 comments
3 Comments:
Ever heard that if you put a frog in hot water, it jumps out, but if you put a frog in cold water, then very slowly bring the water to a boil the frog will never notice the gradual increase in temperature and sit there gazing at the world around it even as it boils to death?
My department too is talking about scanning literally tons of paper documents dating back to the early part of the last century. The documents are, of course, completely useless now but that's beside the point. We're well into the 21st century and apparently it never occured to anyone even ten years ago that scanning these documents might be a good idea. We're still just "tossing the idea around" even as our paper documents continue to be stuffed, jammed, and forced every which way into the seemingly endless line of filing cabinets.
Of course, the decision to go for it is also being held back (and I'm serious about this) by the lack of qualified people to scan documents. We've got maybe 4 or 5 people on our team of 19 who are definitely over qualified and thus useful for doing our real work. The others are certainly expendable, unfortunately they are so inept that there is concern among some that they cannot be trusted to consistently scan, name and electronically file documents.
[steps off soapbox before it gets out of hand]
India. OCR software only gives you 90-95% accuracy on the clearest of documents. Sending stuff overseas to be keyed in can give you 99.999% accuracy at relatively affordable costs. Of course, when you toss in the overhead costs of government bureaucracy...
lol...all that sounds like a job for B.S.
Post a Comment
<< Home