A wordfilter (sometimes referred to as just "filter" or "censor") is a script typically used on Internet forums or chat rooms that automatically scans users' posts or comments as they are submitted and automatically changes or censors particular words or phrases.
The most primitive wordfilters search only for a specific string and replaces it regardless of the situation. More advanced wordfilters will make distinctions against certain words. The most advanced wordfilters may use regular expressions.
Functions of wordfilters
Wordfilters can serve any of a number of functions.
Removal of vulgar language
A swear filter, also known as a profanity filter or language filter is a software subsystem which modifies text to remove words deemed offensive by the administrator or community of an online forum. Swear filters are common in custom-programmed chat rooms and online video games, primarily MMORPGs. This is not to be confused with content filtering, which is usually built into internet browsing programs by third-party developers to filter or block specific websites or types of websites. Swear filters are usually created or implemented by the developers of the Internet service.
Most commonly, wordfilters are used to censor language considered inappropriate by the operators of the forum or chat room.
Some swear filters do a simple search for a string. Others have measures that ignore whitespace, and still others go as far as ignoring all non-alphanumeric characters and then filtering the plain text. This means that if the word "you" was set to be filtered, "y o u" or "y.o!u" would also be filtered.
Clichés—particular words or phrases constantly reused in posts, also known as "memes"—often develop on forums. Some users find that these clichés add to the fun, but other users find them tedious, especially when overused. Administrators may configure the wordfilter to replace the annoying cliché with a more embarrassing phrase, or remove it altogether.
Internet forums are sometimes attacked by vandals who try to fill the forum with repeated nonsense messages, or by spammers who try to insert links to their commercial web sites. The site's wordfilter may be configured to remove the nonsense text used by the vandals, or to remove all links to particular websites from posts.
Lameness filters are text-based wordfilters used by Slash-based websites to stop junk comments from being posted in response to stories. Some of the things they are designed to filter include:
- Too many capital letters
- Too much repetition
- ASCII art
- Comments which are too short or long
- Use of HTML tags that try to break web pages
- Comment titles consisting solely of "first post"
- Any occurrence of the word "gay" or other terms deemed (by the programmers) to be offensive/vulgar
Since wordfilters are automated and look only for particular sequences of characters, users aware of the filters will sometimes try to circumvent them by changing their lettering just enough to avoid the filters.substitutions; others may make filter evasion a punishable offense of its own.leet. More advanced techniques of wordfilter evasion include the use of images, using hidden tags or Cyrillic characters. Another method is to use a soft hyphen. A soft hyphen is only used to indicate where a word can be split when breaking text lines and is not displayed. By placing this halfway in a word, the word gets broken up and will in some cases not be recognised by the wordfilter.
Some more advanced filters, such as those in the online game RuneScape, can detect bypassing. However, the downside of sensitive wordfilters is that legitimate phrases get filtered out as well.
Wordfilters are coded into the Internet forums or chat rooms, and operate only on material submitted to the forum or chat room in question. This distinguishes wordfilters from content-control software, which is typically installed on an end user's PC or computer network, and which can filter all Internet content sent to or from the PC or network in question. Since wordfilters alter a user's words without his or her consent, some users still consider them to be censorship, while others consider them an acceptable part of a forum operator's right to control the contents of the forum.
A common quirk with wordfilters, often considered either comical or annoying by users, is that they often affect words that are not intended to be filtered. Some words that have been filtered accidentally can become replacements for profane words. One example of this is found on the Myst forum Mystcommunity. There, the word 'manuscript' was accidentally censored for containing the word 'anus', which resulted in 'm****cript'. The word was adopted as a replacement swear and carried over when the forum moved, and many substitutes, such as " 'scripting ", are used (though mostly by the older community members).
Place names may be filtered out unintentionally due to containing portions of swear words. In the early years of the internet, the British place name Penistone was often filtered out from spam and swear filters.
Many games, such as World of Warcraft, and more recently, Habbo Hotel and RuneScape allow the user to turn the filters off. Other games, especially free Massively multiplayer online games, such as Knight Online do not have such an option.
Other games such as Medal of Honor and Call of Duty (except Call of Duty: World at War, Call of Duty: Black Ops, and Call of Duty: Black Ops 2) do not give users the option to turn off scripted foul language, while Gears of War does.
- Internet censorship
- Content-control software
- Scunthorpe problem
- Filtering of Google Search suggestions
- Online Text Obfuscator — replaces characters with similalar Unicode chars from different character sets (e.g. Cyrillic)
- Webcensor — profanity filter for Google Chrome browser.
- Sheerin, Jude (29 March 2010). "How spam filters dictated Canadian magazine's fate". BBC Online. Retrieved 5 April 2011.