You may wonder what the difference between a FAQ chatbot and a standard built conversation tree flow on Stella is. Imagine you have 100 FAQs for your chatbot. If you build according to the standard procedure, you would need 100 tree nodes to accommodate all the FAQs - it would be a pain to manage!
With just the effort of building a handful of nodes, FAQ module allows you to manage all your FAQs in the data source. You can manage unlimited FAQ entries in Excel and import it as a .CSV file to the data source. Then, you can let Stella do the magic.
Level One - Build FAQ Chatbot Tree with Exact Keyword Match
This section will guide you through the process of building a basic FAQ chatbot that matches exactly the user input with the keywords in the data source.
Level Two - Build FAQ Chatbot with Exact Keyword Match & Diversion
After you have completed level one, you can also add in the diversion feature. This feature allows the display of multiple FAQ entries if there are multiple keyword groups sharing the same keyword in the data source.
Level Three - Build FAQ Chatbot with Keyword Groups Match & Diversion
Instead of matching with the exact keyword, this feature allows Stella to match the entries by the highest score, i.e. the number of keyword groups successfully matched.
If a user typed a question which includes only parts of the keyword groups, the chatbot will still be able to search for an entry and reply with an answer.
Level Four - Advanced Function & Analytics
You can actually trigger any existing chatbot tree flow with the help of FAQ data source and also add analytics to display the total & unique count of each matched FAQ entry.
- Redirect to Existing Chatbot Tree
- Data Analytics for FAQ Chatbot
- Natural Language Processing Integration
Level Five - Filtering Questions
Apart from typing the text directly to the FAQ chatbot, filtering questions allows users to find their answer in a pre-set flow (similar to IVRS in phones). You can categorize your FAQ entries into 3 (or more) layers, so that users can just select the categories layer by layer until they reach the final answer.