The latest news, trends and information to help you with your recruiting efforts.

Posted August 27, 2010 by Adds Language Filters

It has been an interesting and education process this past month as we continue to make the move from a Web 1.0 platform where virtually all communication was from our staff and our employer clients to the job seekers over to our Web 2.0 platform where far more communication is user generated. We knew going into this, for example, that we’d inevitably have some users who wouldn’t post appropriate content and that some of that content may end up offending other users. When I outlined my vision for what we were and where I wanted to take us, I got a lot of “are you crazy” looks and remarks back from friends, family, employees, contractors, clients, and other stakeholders. One of their biggest concerns was how would we deal with the inevitable person who didn’t just accidentally step over the line but deliberately leaped over the line by posting porn or other such content which would be very offensive and potentially harmful to our business.

Fortunately, we’ve had only one real problem in the month but I’m realistic enough to know that we’ll have more and probably more frequent. We’ve been able to minimize the number of incidents by being proactive and responsive. On the proactive side, only members can post content and all members must verify they’re for real by registering at our site, receiving an email to the email address they provide, clicking on a link within that email, completing their profile, and answering a math question to help verify that they’re actually human rather than a “bot” (computer program). Are those hoops fool proof? Of course not, but neither is the lock on your front door but locking your house makes it far less likely that you’ll be robbed so you lock it even though you know that a very determined burglar will smash down your door, break a window, whatever.

So what do we do to prevent illicit content once the member is registered? We moderate all blog, video, photo, and other such content. Unless you’re one of the administrators, if you post a blog article then that blog article does not go live until we’ve reviewed and approved it. There are a handful of blog contributors who post frequently and their content is always top notch so we don’t even have to read their blogs. We just approve them. But if we don’t know you well, we’re going to read your blog article to make sure that it isn’t objectionable. It may not be beautiful prose, but we’ll approve it if it isn’t terribly objectionable.

Over time, I anticipate that we’ll see more and more blogs, videos, and other such content posted to our site and we’ll welcome that. So the review process will become more and more of a burden. How will we minimize the burden on our staff time while still doing what we can to minimize the illicit content? One way is our new language filter. This administrative feature enables us to filter out offensive or otherwise prohibited words from content on which will help us more efficiently moderate the user generated content, as well as increase the value of your site traffic to the rest of our members and clients. The language filter is sort of a 2010 version of George Carlin’s “seven words you can’t say on the radio” skit where he said aloud seven of the most popular swear words. Maybe it is a reflection on society, but our filter has a lot more than seven words in it. In fact, it has dozens upon dozens.

So what happens if you post a blog article with one of the words in it? Chances are that you’re a “real” person and that you just stepped over the line. Maybe you did so accidentally and maybe you did so deliberately but if your blog article is good other than a four letter word that you included, we’ll want to publish your blog article but not with the swear word. So the language filter automatically edits the swear word so only the first letter of the swear word is shown and the rest are replaced with (*). So if “swear” was a banned word and you entered “swear” in your blog article, the word will will show up in your blog article as “s****” (without the quotes). Words are filtered when the page is being prepared for display to the user.

Neat, huh?