Guide FloatChat

Floatchat- User Guid

Enabling Multiple Languages

  1. Introduction #

Floatchat offers a range of capabilities to support the development and management of multilingual bots. These capabilities cover key components such as FAQs, Intents, and conversational flows. Here are some key aspects of these capabilities:

Tagging FAQs for NLP: You can tag FAQs and Intents with specific languages to better manage them and utilize language-specific NLP models. User queries are also resolved with language considerations in mind.

Localization of Messages in Conversation Flows: Instead of building separate paths for different languages, you can build a single path and add strings for the different languages you support.

Localization of Settings Messages: Settings and configurations can also be managed through language properties files. This includes the bot’s name, configuration messages (e.g., user feedback), and other fixed messages. This allows for customization and different messages in different languages.

Language-specific Synonyms and Stopwords: You can define separate sets of synonyms and stopwords for different languages, which will be appropriately utilized by the system.Live Chat Setting

  1. Using Localization in Floatchat #

To utilize these localization capabilities, let’s create a simple chatbot that welcomes users and allows them to explore the chatbot in their preferred language, using English and Arabic as examples.

2.1 Availability #

Multilingual capability is available in our Professional and higher plans.

2.2 Enabling Localization for Your Chatbot #

To enable this feature, go to the “Configure” section in the left-hand menu, select “Languages,” and click on “Enable.” This sets up properties files for managing user-facing messages in paths and supports multilingual functionality.

After enabling this feature, you can see the available languages and manage the strings for various paths and settings in this section. The default language initially set is English, but you can change it later if needed.

All set. Activate your bot to get started. (1)

2.3 Managing Localized Messages with Properties Files #

All messages from the Conversation Flow Modeler and Bot Configurations are identified and referred to as keys. You can download a properties file containing all the keys and corresponding values for easy message management and translation. Ensure that you include the translated messages corresponding to each key/line in the downloaded file. Upload the key file for each new language. It’s essential to avoid missing any messages/strings in the properties files.

2.4 FAQs and Intents #

FAQs can be associated with specific languages, allowing you to have multiple sets of FAQs corresponding to the languages you support. To add FAQs for your bot:

– Go to the FAQ workflow from the Build section.

– Click on “ADD FAQ” to add queries and their variations.

– Select the language associated with the FAQ set.

– Enter the queries and their responses, and click on “ADD” to add and tag the FAQ to the language.

When a user asks a query, the system will look up FAQs stored in the user’s selected language and, additionally, any FAQs associated with the bot’s default language.

2.5 Control the User Flow #

To change the language or determine and set the user’s language, you can use the “Select Language” node in the path. This node allows the user to switch to their desired language, and all subsequent nodes/paths will be displayed in that language. The NLP engine will use the set language to determine the response.

There are alternative methods for setting the language, such as via the website chat widget script or using attributes in nodes like Set Attributes or JSON API nodes.

2.6 Default Language

The default language is the language initially displayed by your bot. By default, it is set to English but can be changed. The default language serves as a fallback when specific language strings are missing. You can change the default language, either by associating existing FAQs and copying missing strings or by keeping them as they are.

  1. Best Practices for Building Multilingual #

 Bots

Here are some best practices to consider when building multilingual bots:

3.1 Easily Change the Language #

Ensure that it is easy for bot users to find options for changing languages. Use the Select Language node or control language settings via attributes, depending on user configurations.

3.2 Include Multi-language Instructions #

Add messages for changing the language in all supported languages. Also, include introductory messages and common keywords/actions in all languages for a seamless experience.

3.3 Avoid Missing Values #

Review all strings/messages in all supported languages to check for missing values and ensure accuracy, including settings and default values. Users should not see keys instead of messages.

3.4 Avoid Using Special Characters #

Special characters like *, “”, ^.$, and # may cause errors for some languages. Avoid using them in keys and attributes.

3.5 Language Not Supported #

If a specific language is not directly supported, choose the closest language and add the necessary FAQs and path messages. English can also be used, relying on the default matching rules engine.

3.6 Supported Languages #

Floatchat currently supports 52 languages, including Arabic, Bengali, Chinese, English, French, German, Hindi, Italian, Japanese, Korean, Russian, Spanish, and many others.

References

– NLP: Learn more about training your bot using Floatchat’s proprietary NLP Engine.

– Conversation Flow Modeler: Learn more about building conversation flows.

– Synonyms & Stopwords: Learn more about managing synonyms and stopwords for your bot.

Note: The content has been rewritten for Floatchat, but it’s important to note that Floatchat is a fictional application created for the purpose of this exercise, and the specific features mentioned may not exist in any real-world software or product.

 

What are your feelings
Scroll to Top