RASA – an Open Source Chatbot Solution

Conversational agents or chatbots are the current trends which is followed by the organizations as their first line of support for reducing the response time. The initial generation of bots was not very smart as they understood only a limited set of queries.

Different Chatbot platforms are:

-        DialogFlow vs.  RASA

DialogFlow:

• -        Formerly known as API
• -        It does almost everything what chatbots generally do
• -        Entities and intents classification is handled by DialogFlow
• -        Lacks end-user management
• -        It allows webhooks for fulfillment
• -        Its robust API allows defining intents and entities
• -        Hosts data in clouds
• -        On-premise operation is not possible

RASA is an open-source machine learning the tool used by developers and product teams for expanding bots ability to answer more than just simple questions.

RASA’s Terminologies:

-         Intent: It is the basic intention or the purpose of the input by the user

-        
Entity: This is using the information which is derived from the user input which can be extracted like time or place
    
Actions: It is an operation which is performed by the bots

Stories: These are simply the conversations between bots and the user

RASA Stack:

RASA has two basic components. They are:

-        

      RASA NLU: It is a library for understanding natural language which provides the basic functions of the entity extraction and intense classification. The users are understood by the chatbots through this.

-        RASA Core: Here the dialog flow of the system is generalized through Machine Learning techniques.

RASA Architecture: The steps are:

1. -        The received message is passed to the interpreter
2. -        The interpreter converts it into a dictionary which includes the original text, entities, and intents which are found
3. -        The tracker keeps a track of the conversations
4. -        The current state of the tracker is received by the policy
5. -        Action to take the next is done by policy
6. -        The tracker logs the chosen action
7. -        A response is sent to the user

RASA can be used for:

• -        Customer service
• -        Financial service
• -        Healthcare

GAVS Technologies is a leading Managed Security Services Provider in USA who provides the Digital Service Desk AI Software.