{"id":18016,"date":"2017-12-26T09:51:06","date_gmt":"2017-12-26T09:51:06","guid":{"rendered":"http:\/\/developer.kore.ai\/v10-1\/?post_type=docs&p=12311"},"modified":"2024-02-13T06:24:32","modified_gmt":"2024-02-13T06:24:32","slug":"building-multi-language-bots","status":"publish","type":"post","link":"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/advanced-topics\/multi-lingual\/building-multi-language-bots\/","title":{"rendered":"Getting Started with Multilingual Virtual Assistants"},"content":{"rendered":"

Consumers are more likely to engage with virtual assistants that communicate in their preferred language. The Kore.ai XO Platform supports enabling multiple languages within an assistant without having to rebuild the definitions. The platform supports over 100 languages and you can choose to enable any of these languages for your assistant. You can start building with one language and enable additional languages as and when you need them.\u00a0\u00a0<\/span><\/p>\n

This article takes you through the general context within which multilingual VAs work. It discusses language use in VA-user conversations and in NLU processes, language enablement options, as well as language detection and selection.\u00a0<\/span><\/p>\n

<\/div><\/div>\n

Building a Multilingual VA<\/span><\/span><\/h1>\n

If you want to build a Multilingual VA, there are a few points to keep in mind:<\/span><\/p>\n

    \n
  1. There is a set of basic building blocks to a multilingual VA: the language in which it talks to users, the language in which you train it, and the process through which it detects and selects which languages to use. Please continue reading this article to learn more.<\/span><\/li>\n
  2. You can create a new Virtual Assistant or add new languages to an existing one. Read more about managing languages <\/span>here<\/span><\/a>.<\/span><\/li>\n
  3. Multilingual VAs have language-specific components and some features exhibit different behaviors compared to single language assistants. Read more about this <\/span>here<\/span><\/a>.<\/span><\/li>\n
  4. Translation can be automated using pre-built translation services from providers such as Microsoft or Google; or custom services, including any that you may build in-house. Read here<\/a> to learn more.<\/span><\/li>\n<\/ol>\n
    <\/div><\/div>\n

    The Conversation (Bot) Language<\/span><\/span><\/h1>\n

    Enabling a language requires you to train the model to understand the user\u2019s input and present the responses in the user\u2019s language. To achieve this, the platform allows you to choose a <\/span>Conversation (Bot) Language<\/strong> and <\/span>NLU Language<\/strong> for every language that you would like to enable. In most cases, the Conversation Language can be the same as the NLU Language.<\/span><\/p>\n

    Conversation Language is the one that users use to interact with the assistant. You can choose any of the over 100 languages as the Conversation Language. You can define the responses (prompts, messages. etc.,) in the Conversation Language.\u00a0<\/span><\/p>\n

    You can also use the automatic response translation feature when you or your team does not have expertise in the user\u2019s language. You can write the responses in your preferred language and the platform will automatically translate them to the user\u2019s language during the conversation.\u00a0<\/span><\/p>\n

    <\/div><\/div>\n

    Supported Bot Languages<\/strong><\/span><\/h2>\n

    The following are the Conversation Languages supported by the Platform:<\/p>\n

     <\/p>\n

     <\/p>\n

     <\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
    Afrikaans \u2013 af<\/span><\/td>\nHaitian_creole \u2013 ht<\/span><\/td>\nRomanian \u2013 ro<\/span><\/td>\n<\/tr>\n
    Amharic\u00a0 \u2013 am<\/span><\/td>\nHungarian \u2013 hu<\/span><\/td>\nRussian \u2013 ru<\/span><\/td>\n<\/tr>\n
    Assamese \u2013 as<\/span><\/td>\nIrish \u2013 ga<\/span><\/td>\nSinhalese \u2013 si<\/span><\/td>\n<\/tr>\n
    Arabic \u2013 ar<\/span><\/td>\nIndonesian \u2013 id<\/span><\/td>\nSlovak \u2013 sk<\/span><\/td>\n<\/tr>\n
    Azerbaijani \u2013 az<\/span><\/td>\nIgbo \u2013\u00a0 ig<\/span><\/td>\nSlovenian \u2013 sl<\/span><\/td>\n<\/tr>\n
    Armenian \u2013 hy<\/span><\/td>\nIcelandic \u2013 is<\/span><\/td>\nSpanish \u2013 es<\/span><\/td>\n<\/tr>\n
    Albanian\u00a0 \u2013 sq<\/span><\/td>\nItalian \u2013 if<\/span><\/td>\nSamoan \u2013 sm<\/span><\/td>\n<\/tr>\n
    Bulgarian \u2013 bg<\/span><\/td>\nJapanese \u2013 ja<\/span><\/td>\nShona \u2013 sn<\/span><\/td>\n<\/tr>\n
    Belarusian \u2013 be<\/span><\/td>\nJavanese \u2013 jv<\/span><\/td>\nSomali \u2013 so<\/span><\/td>\n<\/tr>\n
    Bengali\/Bangla \u2013 bn<\/span><\/td>\nKazakh \u2013 kk<\/span><\/td>\nSerbian \u2013 sr<\/span><\/td>\n<\/tr>\n
    Basque \u2013 eu<\/span><\/td>\nKhmer \u2013 km<\/span><\/td>\nSesotho \u2013 st<\/span><\/td>\n<\/tr>\n
    Bosnian \u2013 bs<\/span><\/td>\nKannada \u2013 kn<\/span><\/td>\nSundanese \u2013 su<\/span><\/td>\n<\/tr>\n
    Burmese \u2013 my<\/span><\/td>\nKorean \u2013 ko<\/span><\/td>\nSwedish \u2013 sv<\/span><\/td>\n<\/tr>\n
    Cebuano \u2013 ceb<\/span><\/td>\nKurdish \u2013 km<\/span><\/td>\nSwahili \u2013 sw<\/span><\/td>\n<\/tr>\n
    Catalan \u2013 ca<\/span><\/td>\nKyrgyz \u2013 ky<\/span><\/td>\nTamil \u2013 ta<\/span><\/td>\n<\/tr>\n
    Chinese \u2013 izh<\/span><\/td>\nKinyarwanda -rw<\/span><\/td>\nTagalog – tl<\/span><\/td>\n<\/tr>\n
    Corsican \u2013 co<\/span><\/td>\nLatin \u2013 la<\/span><\/td>\nTibetan \u2013 bo<\/span><\/td>\n<\/tr>\n
    Croatian \u2013 hr<\/span><\/td>\nLuxembourgish \u2013 lb\u00a0<\/span><\/td>\nTelugu \u2013 te<\/span><\/td>\n<\/tr>\n
    Czech \u2013 cs<\/span><\/td>\nLaothian\/Laos\/lao \u2013 lo<\/span><\/td>\nTajik \u2013 tg<\/span><\/td>\n<\/tr>\n
    Danish \u2013 da<\/span><\/td>\nLithuanian \u2013 lt<\/span><\/td>\nThai \u2013 th<\/span><\/td>\n<\/tr>\n
    Dutch \u2013 nl<\/span><\/td>\nLatvian \u2013 lv<\/span><\/td>\nTurkmen \u2013 tk<\/span><\/td>\n<\/tr>\n
    English \u2013 en<\/span><\/td>\nMarathi \u2013 mr<\/span><\/td>\nTagalog\/Filipino – fil<\/span><\/td>\n<\/tr>\n
    Esperanto \u2013 eo<\/span><\/td>\nMalay \u2013 ms<\/span><\/td>\nTurkish \u2013 tr<\/span><\/td>\n<\/tr>\n
    Estonian \u2013 et<\/span><\/td>\nMalagasy \u2013 mg<\/span><\/td>\nTatar \u2013 tt<\/span><\/td>\n<\/tr>\n
    Finnish \u2013 fi<\/span><\/td>\nMaori \u2013 mi<\/span><\/td>\nUighur\/Uyghur \u2013 ug<\/span><\/td>\n<\/tr>\n
    French \u2013 fr<\/span><\/td>\nMacedonian \u2013 mk<\/span><\/td>\nUrdu \u2013 ur<\/span><\/td>\n<\/tr>\n
    Frisian \u2013 fy<\/span><\/td>\nMaltese \u2013 mt<\/span><\/td>\nUkrainian \u2013 uk<\/span><\/td>\n<\/tr>\n
    German \u2013 de<\/span><\/td>\nMalayalam \u2013 ml<\/span><\/td>\nUzbek \u2013 uz<\/span><\/td>\n<\/tr>\n
    Greek \u2013 el<\/span><\/td>\nMongolian \u2013 mn<\/span><\/td>\nVietnamese \u2013 vi<\/span><\/td>\n<\/tr>\n
    Galician \u2013 gl<\/span><\/td>\nNepali \u2013 ne<\/span><\/td>\nWolof \u2013 wo<\/span><\/td>\n<\/tr>\n
    Georgian \u2013 ka<\/span><\/td>\nNorwegian \u2013 nb<\/span><\/td>\nWelsh \u2013 cy<\/span><\/td>\n<\/tr>\n
    Gujarati \u2013 gu<\/span><\/td>\nNyanja \u2013 ny<\/span><\/td>\nXhosa \u2013 xh<\/span><\/td>\n<\/tr>\n
    Hausa \u2013 ha<\/span><\/td>\nOriya\/Odia \u2013 or\u00a0<\/span><\/td>\nYiddish \u2013 yi<\/span><\/td>\n<\/tr>\n
    Hawaiian \u2013 haw<\/span><\/td>\nPunjabi \u2013 pa<\/span><\/td>\nYoruba \u2013 yo<\/span><\/td>\n<\/tr>\n
    Hebrew \u2013 he<\/span><\/td>\nPolish \u2013 pl<\/span><\/td>\nZulu \u2013 zo<\/span><\/td>\n<\/tr>\n
    Hindi \u2013 hi<\/span><\/td>\nPortuguese (Brazilian) \u2013 pt<\/span><\/td>\n<\/td>\n<\/tr>\n
    Hmong \u2013 hmn<\/span><\/td>\nPortuguese (European) – pt_pt<\/span><\/td>\n<\/td>\n<\/tr>\n
    Hmong \u2013 hmn<\/span><\/td>\nPersian \u2013 fa<\/span><\/td>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
    <\/div><\/div>\n

    The NLU Language<\/span><\/span><\/h1>\n

    The NLU Language is the one that you train the assistant with, to identify the user\u2019s intents. The NLU model is built using the NLU Language that you choose. This language can be the same as the Conversation language or it can be any other supported language.\u00a0<\/span><\/p>\n

    <\/div><\/div>\n

    Supported NLU Languages<\/span><\/span><\/h2>\n

    \\The following are the NLU Languages supported by the platform. While most of the NLU features are supported in all languages, there are some exceptions, <\/span>see here for more details<\/span><\/a>.<\/span><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
    Arabic \u2013 ar<\/span><\/td>\nKazakh (post v7.2 release) \u2013 kk<\/span><\/td>\n<\/tr>\n
    Chinese Simplified \u2013 zh_cn<\/span><\/td>\nMarathi (post v9.0 release) \u2013 mr<\/span><\/td>\n<\/tr>\n
    Chinese Traditional -zh_tw<\/span><\/td>\nNorwegian (post v8.1 release) \u2013 nb<\/span><\/td>\n<\/tr>\n
    Catalan (post v9.0 release) \u2013 ca<\/span><\/td>\nPolish (post v7.0 release) \u2013 pl<\/span><\/td>\n<\/tr>\n
    Dutch \u2013 nl<\/span><\/td>\nPortuguese (Brazilian) \u2013 pt<\/span><\/td>\n<\/tr>\n
    English \u2013 en<\/span><\/td>\nPortuguese (European) – pt_pt<\/span><\/td>\n<\/tr>\n
    French \u2013 fr<\/span><\/td>\nRussian (post v7.0 release) \u2013 ru<\/span><\/td>\n<\/tr>\n
    Finnish (post v6.4 release) \u2013 fi<\/span><\/td>\nSwedish (post v7.1 release) \u2013 sv<\/span><\/td>\n<\/tr>\n
    Hindi (post v8.1 release) \u2013 hi<\/span><\/td>\nSlovenian \u2013 sl\u00a0<\/span><\/td>\n<\/tr>\n
    German \u2013 de<\/span><\/td>\nSpanish \u2013 es<\/span><\/td>\n<\/tr>\n
    Indonesian \u2013 id<\/span><\/td>\nTagalog – tl<\/td>\n<\/tr>\n
    Italian \u2013 it<\/span><\/td>\nTelugu (post v9.0 release) \u2013 te<\/span><\/td>\n<\/tr>\n
    Japanese \u2013 ja<\/span><\/td>\nTamil (post v9.0 release) \u2013 ta<\/span><\/td>\n<\/tr>\n
    Korean \u2013 ko<\/span><\/td>\nUkrainian (post v7.0 release) \u2013 uk<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

     <\/p>\n

    <\/div><\/div>\n

    Language-specific NLU Models\u00a0<\/span><\/span><\/h2>\n

    The platform supports language-specific NLU models for 26 languages. These models are pre-trained to understand system entities, concepts, sentiment, etc. in specific languages.\u00a0<\/span><\/p>\n