This is a game targetted at 12 and it would be irresponsible not to make an attempt to keep the language clean.

Filtering is done client side so can turn it off if you want.

First rule of game design: know your audience.

Quote:

Original post by Telamon This is a game targetted at 12 and it would be irresponsible not to make an attempt to keep the language clean. Filtering is done client side so can turn it off if you want. First rule of game design: know your audience.

Any twelve year old, already knows all the bad words backwards both pronouncination and spelling. Most seven years olds speak fluent in ghetto/leet.

An oddity: what if a user's name was "shy-theed" spelt phonetically.

I have faced this oddity. I was playing Runescape, naming myself TheOddMan (obviously). Unfortunately whenever anyone types my name it gets filtered, thanks to the last four letters being close enough to 'damn'.

First, sit down and ask yourself this question.
-What is communication?
Communication is the exchange of feelings, opinions, ideas, questions and facts.
Thus, communication requires a medium which is understood by all parts and agreed upon, forming a dialogue.
As with most if not all the worlds languages, the written ones are also changing. Words which were 'slang' ten years ago, are now a part of most people's vocabulary and accepted as a part of the official language, and included in dictionaries.
Now, what happens if some words, combination of words or combination of characters forming a word is filtered out, and prohibited? Note what I just stated above; communication isn't limited by the words themselves, but the implication the word has, and the assertion that it 'means the same' for the sender and receiver. Slang and new words are pretty widespread in small communities and subcultures. This can be caused by several reasons. For example, the new word might define something which is hard to describe in the group's mother tongue. Or perhaps the implication of the word doesn't have a strict definition in any language. And last, new slang can be made to make the group feel exclusive, and to render the meaning incomprehensible to people outside the group.

This is why simply banning words and character combinations won't work in the long run. Gamers in a particular game ARE a small community, and they will soon agree upon totally new words which implies the exact same meaning, and does not even necessary look like the normal word.
Now, the questions are:

* Do you want to censor the implication of a word, or the word as it is written?
* What group do you base your profanity filtering on? The parents, grown-ups, or the players themselves? Would mothers for example understand what the word "hax" or "n00b" implies?

-Madsy

[Edited by - Madsy on August 17, 2006 6:05:31 PM]

Quote:

Original post by Telamon Our online kids game has recently started to pick up a good number of users and yesterday we got our first email from a mother complaining about some profanity used in the in-game chat. I feel like this is a milestone of some sort, but now we need to do something about it. An obvious first pass would store a dictionary of dirty words in a hashtable and strikeout matching words in the chat text (possibly a word at a time, or possibly using a "sliding window") However, this makes it easy to get around the filter by adding spurious letters ect. Recognizing that there is no way to analyze chat strings for meaning, what is the best approach to solving this problem?

[opinion]
This is the first email, why get so worried about one. Maybe if you get a few hundred then you may think of doing something.
I agree with your point of knowing your audience and from that you should know no matter what you do they will find a way round it, including creating words. If the game was using profanity then I would say you have to do something about it, but think about it many many children use msn and is there a swear filter there?

What next are you going to correct there spelling? :)
[/opinion]

its just not possible to correctly filter all ingame chat if the context of the content being filtered is constantly changing (that is natural language).

you could regex out the obvious profanities, but at the end of the day i feel theres not much point, you probably wont get very far asking your users to kindly not swear either :( such is life

I don't think you can filter everything, but you could match against a wordlist discarding non_characters. You could also convert much used replacement characters to their respective char counterpart. @ -> a, 4 -> a, $ -> s etc and check those as well...

Greetings.