Matching systems

Image of an author
Dariusz Zabrzenski
2 min read
updated: Apr 4, 2022

Every time your client chats with your bot, the system carefully analyses all of the input to match it with the right interaction. You can modify and change the matching system to provide your clients fluent and better user experience.

If you created your bot automatically with ChatBot 2.0, you do not need to plan your story - the bot is created automatically. You can choose which options from your knowledge source users will get while configuring the crawling process. Our AI Assist will handle all queries the bot might not understand.
If you created your bot automatically with ChatBot 2.0, you do not need to plan your story - the bot is created automatically. You can choose which options from your knowledge source users will get while configuring the crawling process. Our AI Assist will handle all queries the bot might not understand.

How Machine System worksLink icon

Matching systems are responsible for pairing user input with User Says field. The system weighs both values and gives the score. If the score is equal or higher than the setup Confidence Score, the bot response is triggered so choosing the right matching systems can be crucial for the seamless conversation flow.

Available Matching SystemsLink icon

To give you better control over your chatbots, we have introduced two matching systems, Machine Learning, and Keywords.

  • Machine Learning uses Natural Language Processing and Algorithmic probability. The system reads the full user input and carefully analyses it. The matching strength depends on the confidence score user setup. ML is the default matching system and it’s automatically enabled.

  • Keywords search for the defined word in the user input. When the system finds the keyword, the matching score is equal to 1 (100%).

Prioritised matching systemLink icon

The ChatBot’s matching system always selects the interaction with the higher matching score. But what about scenarios when the bot must decide between interactions that are ruled by different matching systems?

  • User intents with ML matching are prioritized.

  • When Keywords system find the expected word, the matching score equals 100%.

When a bot meets two interactions with two different matching systems, Machine Learning has preference even if the matching score is lower than the matching score of Keywords.
When a bot meets two interactions with two different matching systems, Machine Learning has preference even if the matching score is lower than the matching score of Keywords.

Check examples and exceptions to see when the above rules won’t be executed.

Was this article helpful?

Got it!

Thanks for your feedback.

Thank you!

We’re happy to help.

Start a free ChatBot trial
and build your first chatbot today!

Free 14-day trial No credit card required

Discover our text| products