There are several stop lists on the admin panel:
The system splits all received text messages with spaces. For example, the user texted “And-after-all, you’re my wonderwall.” The algorithm sees this as a message with four words split with spaces: “And-after-all,” “you’re,” “my,” and “wonderwall.” Each word is checked with the basic rules of illicit language. If the system finds matches, words are restrcited from sending in a chat.
obscene language and rude words, words can be used for sexualized provocations, insults based on nationality and race, threats, insults or specific prohibited topics-words-phrases that are specific to the language, words prohibited in public space by local law.
In addition to the list of words themselves, you need to write down their variations: inflections for nouns, participles, adjectives, conjugations for verbs (if applicable).
It is also necessary for the most popular words to indicate generally accepted spellings (for example, replacing “i” with “1” or “!” and other frequency replacements).
Please remember, moderation works through complete matching: If the list contains the word "panda", the words "pandas" will be allowed.
Words that are homonymous or partially homonymous with valid words (when placing an obscene word inside another word makes it censored). This is especially true for short words-syllables, which will block the sending of censored words. Be careful!
(If you add the word “Mars”, users won’t be able to add “Marseille”).