A challenging debate in social media is the extent to which a brand should censor or empower conversations in which their ‘followers’ express themselves in full technicolour – should people be allowed to use obscenities, swear, incite, or insult on your website?
Profanities: authentic or crude?
It’s a challenging question – should brands allow people to express themselves fully and in so doing potentially offend the sensibilities of someone else and damage their brand; Or should brands cherish authenticity, and freedom of expression, allowing people to use even the most colourful language – and just hope for the best?
Where do you draw the line of acceptability?
There are lots of factors which will influence this; the brand itself, the audience and the domain the conversation is taking place in. Some of it may be impassioned, human and authentic, but more often than not, it’s not. It can be crude, offensive, and brands would simply rather not associate themselves with it. Brands value user generated content (UGC) but more often than not, rightly or wrongly, it’ll be on their own terms – it’ll need to be ‘clean’ to ensure that no one gets offended and kicks up a fuss.
It’s an interesting debate which we’ll cover off in more detail soon, but for now, we’re going to discuss how you can approach automated pre-moderation and swear filters.
Profanity and obscenity filters
So while there’s not a one-size-fits-all approach to censorship, nor a one-size-fits-all solution (other than social media pre moderation), supposing your brand wants to play it safe, one solution is to implement a profanity or swear filter to hide or rewrite obscenities. A simple blacklist won’t catch enough so to improve on this you can implement a fuzzy string search (also called approximate string matching) which looks at the stem of a word to ascertain whether or not it’s likely to be a profanity.
Download a swear list
So it’s all very well creating a fuzzy string search, but what should you search against? Well, being the sharing, caring types, we’ve pulled together a list from various sources of more than 2000 profanities so that you don’t have to! Obviously, whilst this list is extensive, it doesn’t guarantee that someone won’t find a new way to creatively express themselves in a linguistically ‘innovative’ way so unless you’re using a ‘fuzzy’ search as outlined above, you won’t be able to mitigate against the use and display of any word other than those defined in this list.[box type=”download” size=”large” border=”full”]Download Black & White Zebra’s profanity list 2011[/box]
What have we missed?
I expect there may be a few words we missed – help us develop this resource to be even more useful by contributing below!