With Teneo you can create solutions in over 86 languages, varying from Basque to Ukrainian. Additionally, you can use template solutions that will boost your bot with lots of pre-built flows, like greetings and conversational small talk. You can also improve the language understanding of your bot by adding lexical resources, which cover general language vocabulary and common phrases.
Over 86 languages
When you create a solution in Teneo, one of the first things you do is specify its language by selecting it in the Language dropdown of the New Solution window.
Teneo supports the following languages:
- Bengali (Bangla)
- Chinese (Mandarin)
- Kirundi (Rundi)
- Quechuan (Quechua)
- Scottish Gaelic
- Swahili (Kiswahili)
- Tswana (Setswana)
- Zulu (isiZulu)
Based on the language selected, Teneo will use language-specific input processing libraries that handle things like tokenization, sentence splitting, and auto-correction: things that may need to be treated differently per language.
In addition to selecting the language when you create a new solution, you can also choose to base it on a template solution, called Teneo Dialogue Resources. This will boost your bot with lots of pre-built flows. The bot will then, right out of the box, understand and respond to greetings, meta-requests (like Can you repeat that?), feedback (like Great! or I like you), and conversational small talk.
When you request access, your environment will contain dialogue resources for the following languages:
After you have created a solution, you can assign Teneo Lexical Resources to it. Lexical resources are large collections of predefined language objects that can be used in language conditions. The lexical resources cover general language vocabulary and common phrases so that you do not have to model them yourself.
When you request access, your sandbox environment will contain Teneo Lexical Resources covering general language vocabulary for the following languages:
The Lexical Resources are used extensively by the Dialogue Resources mentioned earlier.
Additionally, your sandbox will have lexical resources included in the Lexical Resources to serve specific purposes:
- Sentiment and Intensity
The sentiment and intensity resource contains off-the-shelf building blocks for detecting positive or negative sentiment as well as intensity in user inputs. The resource is available for Chinese, Dutch, English, German, Italian and Swedish.
- Offensive language detector
The offensive language detector resource contains off-the-shelf building blocks for detecting abusive and controversial language in user inputs. The resource is available for English.
- DateTime Handling
The DateTime Handler resource is part of the DateTime Handler extension, which can be used to catch date and time expressions in the user input. This extension is available for Chinese, Danish, Dutch, English, French, German, Italian, Japanese, Norwegian, Portuguese, Spanish and Swedish.
Depending on the language and use case of your bot, you can choose which language resources need to be assigned to your solution.
Creating a non-English solution
In 'Your first solution' you have already seen how you can create a solution that is based on English. There, you chose English as the bot's language because the tutorials on this site assume your bot understands and responds in English.
For a practical example of how to create a non-English solution, please go here: How to create a non-English solution.