Transcribing Comics - A job for the Mechanical Turk?
There are a couple of webcomics I read on a daily basis (a couple a week in the case of MegaTokyo) and recently I’ve found myself wanting to link to a couple of different strips in blog postings; and then discovered that they’re almost impossible to search through.
None of the webcomics I read regularly have any kind of strip content search. You can’t see who was in which strip, you can’t search on the punchlines - which is what I want - and, apart from a couple of sites which have a one sentence summary, you can’t get any more context about any days issue than the time it was uploaded. I understand that transcribing them would be dull as hell but why not just farm the work out via the Mechanical Turk?
The images are already online, you can filter by asking a couple of sample questions (which of these is Dogbert) to determine that the speech will be attributed to the correct person, poll three people per strip to make sure you don’t get bitten by typos, and with a decent “submit results” web form you’d get some nicely formated data (who spoke in which panel for example) for free. And I get to find the Dilbert strip that ends with “You’d have to lift your arms up and run around.”