Design and Implementation of an embodied cognitive conversational agent for assisting in a web site based on the API.AI conversational platform.

Ángel Silván. (2018). Design and Implementation of an embodied cognitive conversational agent for assisting in a web site based on the API.AI conversational platform.. Final Career Project (TFG). Universidad Politécnica de Madrid, ETSI Telecomunicación, Madrid.

Abstract:
This project has focused on the design and implementation of a personal assistant integrated into the web page of the Intelligent Systems Group (GSI) allowing an interaction between the user and the conversational agent through the use of cognitive computing. For this, a chat bot has been developed to carry out the possible conversations. This system has been developed using the Dialogflow alternative as well as a Content Management System called Joomla!. An architecture has been proposed for our system that would enables to perform the planned tasks, integrating a question-answer system, a conversation agent capable of identifying and processing the natural language in Spanish and an information search system on the website where it is integrated the conversational agent. In order that the integrated bot chat was complete, two different ways have been studied according to the user’s request. The first is a light conversation between the user and the bot that enables simple requests without information searches. The second possible way is to abandon this light conversation if questions about member data, publications or GSI projects are detected in the request. In this case, the bot performs a search in the database and provides a summary to the user, if the search has been successful, as well as the possibility of accessing more extensive information. With all this, the project will allow users to engage in a light conversation or search for information through the website and will make it possible to perform exact searches.