This is the example bot used I used in a recent workshop on Rasa.
The presentation discussed the basics of Rasa, starting from what a chatbot is, all the way to building one with Rasa.
This example builds on our v1 chatbot, the code for which you can find here.
- handles the sad path where the user may not want to provide information
- handles the case where the user may provide on partial information, like only their name or only their email
- provides the user with further actions once a valid name and email are supplied
- the user can choose to end the conversation by choosing the "That's all" option
- the user can provide another set of name and email by choosing the "Add More Information" option.
- makes use of custom actions
- makes use of rules
- has mutliple stories to handle different scenarios
A few cases that are handled in v2 but weren't in v1:
The user isn't willing to provide their information
👨 : Hey
🤖 : Please provide your contact information
👨 : don't want to
🤖 : Sorry! You'll have to provide your information to proceed.
The user provides only partial information
👨 : Hey
🤖 : Please provide your contact information
👨 : Sure. The name is John.
🤖 : Thanks. Could you also provide your email?
The user provides no information
👨 : Hey
🤖 : Please provide your contact information
👨 : Sure.
🤖 : Please provide your contact information
The user is shown further actions and chooses to end the conversation
👨 : Hey
🤖 : Please provide your contact information
👨 : Sure. Its john. And email is john@email.com
🤖 : Thanks! What would you like to do next?
- That's all
- Add More Information
👨 : That's all
🤖 : Thanks for your time!
The user is shown further actions and chooses to add more information
👨 : Hey
🤖 : Please provide your contact information
👨 : Sure. Its john. And email is john@email.com
🤖 : Thanks! What would you like to do next?
- That's all
- Add More Information
👨 : Add More Information
🤖 : Please provide your contact information
This project follows the format of a standard Rasa project. There's a directory called data for training data like nlu, stories, and rules.
There's a directory called actions, which contains all your custom actions.
You'll also find the domain.yml file, which mentions all your intents, entities, slots, responses and actions.
Finally, there's the config.yml file, which specifies the components your bot is comprised of.
- Clone this repo
- Navigate to the RasaChatbot directory
- Install rasa>=2.6.2 in an env.
Modify the files in data/ or the domain.yml file to play around.
Before training the bot, a good practice is to check for any inconsistencies in the stories and rules, though in a project this simple, it's unlikely to occur.
$ rasa data validate
To train the bot, we simply use the rasa train command. We'll provide a name to the model for better organization, but it's not necessary.
$ rasa train --fixed-model-name contact_bot_v2
To test your bot, open a new terminal window and start a rasa shell session.
$ rasa shell
This will let you chat with your bot in your terminal. If you want a more interactive UI and a little more debugging information like what intents were identified and what entities were extracted, you can use Rasa X.
You can find me on medium here.