| Summary: | Messaging applications are used on a daily basis for communication between people around the globe. Even though the main purpose of this type of application is to provide a service where people can join in order to interact with one another, it has evolved since and automated conversation has grown. This type of automated conversation is only possible through the implementation of intelligent agents called chatbots. The chatbots objective is to mimic user behavior in order to understand and reply to a given user input. This project aims to further improve the already existing mechanisms for this chatbots so that they can interact with other chatbots so that users can, in a single conversation, have access to multiple chatbots. This is possible through the implementation of a series of features that allow chatbot creators to create groups of chatbots, called federations, and delegate the conversation from one chatbot to another chatbot available in a federation. This approach aims to reduce the number of failed responses and at the same time prevent the user’s frustration because if a chatbot cannot answer, the user needs to find and open a conversation with a chatbot that can answer. Every chatbot has a series of tasks and knowledge that he can use to answer user questions, but there is always a reduced scope so that the chatbot can perform at his best. This solution’s objective is to keep this approach of reduced scopes for better performance, but every chatbot can contribute with his knowledge by lending his services to another chatbot.
|
|---|