Profanity Filter Improvements
Fairplex
This is a follow-up from the canny that was recently closed:
I believe that the root problem has not been resolved. After thinking it over, I have a potential long-term solution to suggest.
In VRChat, we see many bios, avatar/world descriptions, group descriptions, and chatboxes containing offending words.
That said, we also know that the actual profanity filter is not perfect; it flags legitimate words, and many people don't understand why their bios/descriptions get messed and think it's a bug.
So first, I think both the VRC clients and the website need more transparency when this is happening. Let us know when we use forbidden words instead of mixing them together. This could be a simple error message.
Long Term Solution:
VRChat may benefit of an option to toggle bad language visibility, aka a Profanity Filter setting! When enabled, which could be the default state, words that matches the profanity rules get replaced with **** or ♥♥♥♥ or [REDACTED] or something that clearly indicate that this word is not allowed. Ideally, the profanity filter may enforce bios, avatars and worlds descriptions, groups descriptions and even chatbox messages.
When disabled, the profanity filter does not filter anything so bad language can be seen in both VRC clients and VRC website.
Log In
＠BLKZOR｜八黑山｜ㇰㇿ
I find it unbearable that the words I write are being altered without my knowledge. At the very least, users should be given the opportunity to edit their posts themselves before they are published, following notification.
Furthermore, as I understand it, this filtering system appears to be based on English; however, users like myself who do not speak English naturally have no knowledge of English bad words, making it difficult to predict which words will be filtered.
And, of course, in languages other than English, bad words remain unchanged whilst unrelated words are rewritten into gibberish, rendering the filter completely useless and resulting in an unfair user experience. (The post merged into ‘amazing-is-banned-word-in-bio’ addresses this issue in much stronger terms.)