After a brief honeymoon period when we first opened up the web site for non-registered members to post comments, the tidal wave of spam to the site started. You know the stuff, adverts for casinos, viagra, doctoral degrees and such like.

Most of it was caught before being published on the site with the nifty little spam filtering software we run here at Digiplay but even so, deleting 100+ spam messages a day rapidly got a little tiresome. But a solution is at hand...

Rather than closing down posting to non-registered members, we've started to use a neat little system called reCaptcha. What this does is use that little squiggly words - or CAPTCHAS - which are difficult to read for most of us but nearly impossible for computers to fathom.

But the story doesn't end there...

What makes reCAPTCHA particularly neat is not just that it stops the spamming so-and-so from polluting the site but that it takes the answers to the CAPTCHAs that people type in and uses them to improve the quality of electronic versions of public domain classic texts.

Here's how - when texts are scanned and optical character recognition (OCR) is used to translate them into plain text various glitches, misreadings and uncertainties creep in because of the problem that computers have in reading blurry text. What the smart people at reCAPTCHA worked out is that the blurry text which was a problem for digital archivists was a resource for web site owners trying to stop spamming.

So what they did is create a system where each time you fill in a reCAPTCHA form you're actually correcting OCR errors in digital archives. This means that your not just leaving a message on the Digiplay site but you are contributing enriched and developing the collection of publicly available texts provided at The Internet Archive.

You can read more about the reCAPTCHA project at http://recaptcha.net/learnmore.html.