There are two widely used systems of automated censorship:
Which of these is preferable in terms of effectiveness? I'm talking about: Reliably blocking bad content, but not hindering the users in normal usage? Is there another alternative?
I'd say whatever you do, don't do the second item. It'd be better to have it reviewed by a moderator...
There are valid times that swearwords or words that look like swearwords are used:
Ultimately, if you block content, they will just change the content to bypass the filter. They might just do something like this: F * * * (with the actual letters; I'm censoring this here).
I would argue against any kind of automated censorship. If a user uses language that is deemed unacceptable, let the user know so he can change his post.
I find it often helps a lot more to change a users' behaviour when you talk to him, instead of letting technical measures do all the policing for you. In the end, your goal is that people talk to each other with respect, not that certain words aren't used.
Myself, I don't give a barnacle about curse words. I think censors are a bunch of bullcannon for over-sensitive, bilge-drinking halibuts.
That said, when given the ability I actually prefer replacing censored words with humorous alternatives.
For example, an online pirate-themed game I once played replaced certain curse words with rather amusing piratey alternatives:
- ass(es) -> booty(ies)
- asshole* -> barrelstopper*
- *bitch* -> tart*
- blowjob* -> harmonica lesson
- clit -> rosebud
- cock -> mast
- cocksuck* -> bilgedrink*
- cum -> bilge
- cunt* -> shrew
- dildo* -> tickler*
- fag(s) -> landlubber(s)
- *fuck* -> *scupper*
- gay -> happy
- muff -> britches
- *penis* -> *John Thomas*
- pussy* -> kitty
- *shit -> *cannon
- shit* -> barnacle*
- slut -> strumpet
- tits -> bosom
- titty* -> bosom
- twat -> halibut
- whore* -> flirt (previously trollop)
- vagina(s) -> dainty(ies)
Etc. Certain extremely "bad" words were simply not allowed at all. This system worked remarkably well, as users were still able to communicate without much hindrance, and the often-amusing results tended to lighten situations.
That said, this does require a general theme for the community, I do not think a random collection of replacement words would have quite the same value. The trick is to choose a word set that is humorous, consistent, and still allowing users to get their point across.
Simply blocking the posts altogether is always a bad idea. Users will be forced to find a way to post their content and the path of least resistance is to simply come up with creative spellings to work around the filter. By allowing the post but modifying the words afterwards, this creates a barrier out of the extra review and edit steps required for a user to go back and work around the filter - plus the user was not met with the frustrating UX motivation of having to edit just to get their main post out.
Note that, on a much deeper psychological level, this is actually somewhat similar to SE sites' successful system of flags and community edits - it provides constructive, positive feedback on what is acceptable (albeit via direct automated replacements rather than user feedback - but in any case it doesn't e.g. create frustrating UX of not allowing posts or unsatisfying word replacements, which doesn't actually improve long-term culture).
An interesting technique is to not "automate the blocking", since user's will always figure away around it. Also, once they use techniques to hide it, it gets incredibly hard to clean up the mess if moderators leave it alone for a few weeks. Instead, I would fake moderator behavior by not blocking their post or immediately staring it out. instead, clean up entire sections at a time, as if a moderator is doing it. You can use automation1 to do this, just don't make it too frequent. The main advantage of this system is that users will not try to circumvent the system, since they believe its a human doing it.
Having users flag bad posts is also a good addition to the above mentioned system.
When your "censor bot" cleans up a users post, record that. You will soon begin to see certain users writing most of the bad words, you can then block/warn them.
As far as whether b** or ****, I would go for all stars since people really don't want to know what they meant, and showing (part of) it defeats the purpose of censoring in the first place.
1A mix of automation and human decision is best. You can use a bot to flag post for review, or perform the tasks its self. This depends on what you're trying to censor. If a bot does any automated work, be sure to have an appeals process for false positives.
