[{"data":1,"prerenderedAt":131},["ShallowReactive",2],{"content:\u002Fguides\u002Farabic-turkish-hate-speech-moderation":3,"ArticleBody_iN511aINKLRqsxVIkncmyrIgxGq5LT3RmNsPkfCs":126},{"kind":4,"path":5,"locale":6,"title":7,"htmlTitle":8,"description":9,"navTitle":10,"published":11,"updated":11,"alternates":12,"breadcrumbs":13,"indexable":21,"twin":22,"collection":23,"slug":24,"file":25,"content":26,"hub":19,"titleEm":28,"eyebrow":29,"lead":30,"cta":31,"faq":34,"faqJsonLd":53,"sources":60,"related":101,"topic":122,"jsonLdType":123,"blocks":124,"shortDefinition":125,"feed":21},"article","\u002Fguides\u002Farabic-turkish-hate-speech-moderation","en","How to moderate hate speech and insults in Arabic and Turkish","Arabic & Turkish Hate Speech Moderation · Detectivision","Why English-first moderation misses Arabic and Turkish abuse, which APIs list both languages for which categories, and how to test one on your own messages.","Arabic and Turkish moderation","2026-11-05T21:36:00+03:00",[],[14,17,20],{"name":15,"path":16},"Home","\u002F",{"name":18,"path":19},"Guides","\u002Fguides",{"name":10,"path":5},true,"\u002Fguides\u002Farabic-turkish-hate-speech-moderation.md","guides","arabic-turkish-hate-speech-moderation","content\u002Fen\u002Fguides\u002Farabic-turkish-hate-speech-moderation.md",{"collection":27,"path":5},"guides_en","Arabic and Turkish","Guide","Many moderation APIs were built for English first. Some list Arabic for one category only, and some do not list Turkish at all. Check what each API publishes for each language and each category, then test it on your own messages. Detectivision checks Arabic and Turkish text for insults, hate speech, racism, sexism, sexual language and profanity, among 12 languages, with no language to set.",{"label":32,"to":33},"Text moderation API","\u002Ftext-moderation",[35,38,41,44,47,50],{"q":36,"a":37},"Which content moderation APIs support Arabic and Turkish?","As read on 5 November 2026: Detectivision checks both, among 12 languages. Hive lists both for its sexual, violence, hate, bullying and phone number classes, under an annual contract. Sightengine lists Turkish. Perspective lists Arabic for TOXICITY only. Azure AI Content Safety's text classifiers are trained and tested on eight languages, neither of them. Amazon Comprehend's toxicity detection supports only English.",{"q":39,"a":40},"Does Perspective API support Turkish?","Its Attributes & Languages page, read on 5 November 2026, lists no attribute in Turkish, and lists Arabic only for TOXICITY. Perspective API also stops after 31 December 2026.",{"q":42,"a":43},"Why do word lists miss Turkish insults?","Turkish builds words with suffixes, so one insult has dozens of forms. A list of whole words misses them; a list that matches inside words blocks harmless ones. People also type without Turkish letters, so ş becomes s and ı becomes i.",{"q":45,"a":46},"Do I need to tell the API which language a message is in?","Not with Detectivision. Send the message as it was written, Arabic or Turkish script and all; there is no language setting, and nothing is translated first.",{"q":48,"a":49},"How do I test a moderation API on Arabic and Turkish messages?","Take a few hundred messages your moderators have already judged in each language, including dialect, Arabic typed in Latin letters, Turkish typed without its letters and quoted abuse. Send them all, then count misses and false alarms for each category before you set your lines.",{"q":51,"a":52},"Can voice messages in Arabic and Turkish be checked too?","Yes. Detectivision's audio moderation understands the same 12 languages as its text moderation, and gives each finding with its start and end in seconds.",[54,55,56,57,58,59],{"q":36,"a":37},{"q":39,"a":40},{"q":42,"a":43},{"q":45,"a":46},{"q":48,"a":49},{"q":51,"a":52},[61,66,69,73,77,81,85,89,93,95,98],{"title":62,"url":63,"publisher":64,"accessed":65},"About the API - Attributes and Languages","https:\u002F\u002Fdevelopers.perspectiveapi.com\u002Fs\u002Fabout-the-api-attributes-and-languages","Jigsaw","2026-11-05",{"title":67,"url":68,"publisher":64,"accessed":65},"Perspective API (sunset notice)","https:\u002F\u002Fperspectiveapi.com\u002F",{"title":70,"url":71,"publisher":72,"accessed":65},"Region availability and service limits: Language considerations (Azure AI Content Safety)","https:\u002F\u002Flearn.microsoft.com\u002Fen-us\u002Fazure\u002Fai-services\u002Fcontent-safety\u002Fregion-availability","Microsoft",{"title":74,"url":75,"publisher":76,"accessed":65},"Trust and safety: Toxicity detection (Amazon Comprehend Developer Guide)","https:\u002F\u002Fdocs.aws.amazon.com\u002Fcomprehend\u002Flatest\u002Fdg\u002Ftrust-safety.html","Amazon Web Services",{"title":78,"url":79,"publisher":80,"accessed":65},"Text Moderation - Overview (languages table)","https:\u002F\u002Fdocs.thehive.ai\u002Fdocs\u002Fclassification-text","Hive",{"title":82,"url":83,"publisher":84,"accessed":65},"What languages does text moderation work with?","https:\u002F\u002Fsightengine.com\u002Ffaq\u002Flanguages-for-text-moderation","Sightengine",{"title":86,"url":87,"publisher":88,"accessed":65},"Moderation (OpenAI API guide)","https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fguides\u002Fmoderation","OpenAI",{"title":90,"url":91,"publisher":92,"accessed":65},"Text moderation (docs)","https:\u002F\u002Fdetectivision.ai\u002Fdocs\u002Ftext","Detectivision",{"title":32,"url":94,"publisher":92,"accessed":65},"https:\u002F\u002Fdetectivision.ai\u002Ftext-moderation",{"title":96,"url":97,"publisher":92,"accessed":65},"Türkçe küfür ve hakaret tespiti (Turkish suffixes, letters and word lists)","https:\u002F\u002Fdetectivision.ai\u002Ftr\u002Fturkce-kufur-ve-hakaret-tespiti",{"title":99,"url":100,"publisher":92,"accessed":65},"Audio moderation API","https:\u002F\u002Fdetectivision.ai\u002Faudio-moderation",[102,106,110,114,118],{"path":103,"title":104,"description":105},"\u002Fturkish-offensive-language-detection-api","Turkish offensive language","Detect insults, hate speech, racism, sexism and sexual harassment in Turkish text, and in English too, with one API call and a confidence per flag.",{"path":107,"title":108,"description":109},"\u002Fhate-speech-detection-api","Hate speech detection","Find hate speech, racism and sexism in text written in 12 languages, and hateful images, with one API call. Each flag comes with how sure it is.",{"path":111,"title":112,"description":113},"\u002Falternatives\u002Fperspective-api","Perspective API alternative","Perspective API shuts down after 31 December 2026. Its attributes mapped to Detectivision flags, the languages each covers, and the code before and after.",{"path":115,"title":116,"description":117},"\u002Fguides\u002Fai-hate-speech-moderation","Can AI moderate text for hate speech and harassment?","Yes, for most of it: a model reads every message and says which kind of abuse it finds and how sure it is. What it does well, where people decide.",{"path":119,"title":120,"description":121},"\u002Fguides\u002Fdetect-toxic-comments","How to detect toxic comments in an app or community","How to detect toxic comments: what counts as toxic, the three approaches, what to do with a score, and a working Python and Node.js example.","by-content","TechArticle",false,null,["Island",127],{"key":128,"result":129},"ArticleBody_iN511aINKLRqsxVIkncmyrIgxGq5LT3RmNsPkfCs",{"head":130},{},1793904542710]