Rasa is a python framework that helps us to build any kind of Chatbot easily. It based on NLU (Natural Language Processing) which offer the possibility to understand what the user want. Let’s discover it.
Firstly you must know how the Chatbot works. The Chatbot is a conversational agent which talk with a client or user and provide answers to the users’ requests independantly of the human intervention.
The Chatbot is composed by some specifics terms that we will discover:
Intents: It represent intentions that the user want to express when he submit his message to the Chatbot. Inside the intents, we provides somes phrases that the user may ask and we also provide some responses that the Chatbot must use to answer to the user. It’s the work of the Developper here.
Entities: Entities are keywords that represent some specifics datas that the Chatbot may use to perform the discussion with the user. Entities are used to extract some values inside the user input (message).
Now, we know how Chatbot works, we can switch to the Rasa Framework.
Rasa is split into two python librairies like Rasa NLU and Rasa Core. The first one is the Natural Language Understanding and use the intents to understand what the user want and entities to extract some specifics values to make the conversation more interesting, after Rasa NLU have understood what the user want and extracted somes values, it send it to the next one. Rasa Core receives the data sended by Rasa NLU and process it to find the correct answer that it should send to the user as output, For having a response, it will look for the responses that the Developper provided to him inside the intents.
1. How Conversational AI can Automate Customer Service
2. Automated vs Live Chats: What will the Future of Customer Service Look Like?
3. Chatbots As Medical Assistants In COVID-19 Pandemic
4. Chatbot Vs. Intelligent Virtual Assistant — What’s the difference & Why Care?
- The User submit the request (a message is sended to the Interpreter)
- In the Interpreter section, we have Rasa NLU which capture the request, and extract intents to understand what the user want, and in the same time it extract some specifics keywords (entities) which represent the important data. Then it send the result to the Tracker. The rest of the elements below is contained inside Rasa Core.
- The Tracker here, is used to store the conversation history in memory. By this fact, the Chatbot should know in where level the conversation is, it maintains the conversation state.
- After that, the Policy comes; it choose the action which will be execute at every step in the dialogue, regarding the history of the conversation.
- The Action here, look inside the history for consulting the state of the conversation. This step can trigge another action.
- The action is executed and an ouput is generated (message from the Chatbot) to the user.
Now you have a brief understanding on how Rasa Framework works, you can now decided to learn about how you can build your own chatbot using this framework.
just check out the documentation of Rasa.
for more explanation: https://rasa.com/docs/rasa/playground
Thank to you reading.