Forty-five percent of employers reported in a 2009 Harris Interactive survey for CareerBuilder that they use social networking sites to research job candidates, with Facebook edging out LinkedIn as the number one social networking site that they check. Ever since those results came out back in August of 2009, there have been many blog posts, presentations, advice columns, and so on advising you to watch what you post, clean up your profile, and cultivate your personal brand. This has also led to a boom in new apps for scrubbing your online reputation. The latest one getting buzz on Mashable.com is socioclean.
My Facebook privacy settings are all set to Friends Only, but with all the advice swirling around us about what things to say or not say on Facebook, Twitter, or blogs many people have gotten quite anxious, including me. Therefore, I signed up with socioclean and ran the app over my Facebook profile recently.
I am very cautious with my postings on Facebook, lately even avoiding references to being sick or tired, never mind sex, drugs, or rock and roll. OK, so I do mention rock and roll once in a great while. Anyway, those who know me in real life know that I don't smoke or drink and am in a long term committed relationship with my partner. I'm pretty boring that way. You probably wouldn't want to invite me to your wild parties. Imagine my surprise when socioclean found stuff that it claimed I needed to clean up.
The tool basically scans wall posts, status messages, photos, and groups for words they deem inappropriate. There is no semantic analysis. If the word appears, it's flagged regardless of context. The results are displayed as a pie chart showing areas of concern and bar charts showing the count of searched words in each category. Clicking on the bar chart gets you to the report showing each searched word that showed up, the complete text, the date, the category, and who created it along with a link to view the offending post. Then you have an option to set the tool to ignore either the word or the instance of the word. Therefore, if something is OK in context, you click ignore instance.
So, the cause of my surprise? Every one of my posts about the Moby Dick marathon reading was tagged as a sexual reference, giving me a percentage of sexual references that earned me a grade of D. If that wasn't humorous enough, a post about the workers removing the ice dam on my roof making banging noises was also tagged sexual and a reference to a glue gun was tagged as aggression. Funniest of all, though, was the puzzling tagging of two references to an article about a night heron rookery in the ornithological journal The Auk as alcohol/drug references. Why? The author was named Waldo Bailey. By the time I excluded every reference to Moby Dick, a few references to airplanes and fishing boats, the glue gun reference, and a friend's comment that a photo was too damn cute, I was left with one genuine alcohol reference to be removed: a tweet about a presentation on locally brewed beers that I heard at Ignite Boston. Hmm, the tool knew enough to exclude Ignite as an aggression reference but not glue gun? Anyway, I deleted the post about the beer presentation, and ran the tool again. I got an A. Clean as a Whistle.
Overall, I don't think the tool was that helpful. I did just as much work reviewing posts as I would have if I had just gone over all my posts manually. The number of false alarms was irritating. I'm still scratching my head about Waldo Bailey. There are a lot of people named Bailey who are not famous Irish beverages.
Semantic issues aside, there are still problems with this tool. It does not take into account your privacy settings, so it sees everything. Therefore, it does not give you a true picture of what an employer or prospective employer would see, assuming Facebook's privacy settings actually work. Socioclean pitches it as a tool that you can use to police yourself. However, what if an employer had your Facebook password? That's not so far-fetched given the recent stories about government agencies requesting prospective employees to turn over their Facebook passwords. If the prospective employer runs the tool and only looks at the graphs, you could lose out on a job because you read Moby Dick or articles by ornithologists named Bailey.
Maybe the real question is not about how offensive your Facebook profile is, but how not to let the constraints of a Puritanical culture in the age of Facebook's philosophy of radical openness stifle genuine human expression.
No comments:
Post a Comment