LIA : um chatbot inteligente para o domínio de imóveis

Abstract

Nowadays the use of intelligent conversational agents to assist in the development and execution of administrative and strategic tasks of companies and organizations has become more and more frequent. There are currently two main types of chatbots: rules-based chatbots and those that use artificial intelligence and machine learning algorithms and techniques. These AI-based agents or chatbots go one step further than predefined rule-based chatbots that already have a predetermined chat flow, and depending on the user’s response, they follow a certain path. In this work, initially a systematic review was developed about which tools, methods and techniques exist in the literature on the subject. Subsequently, market research was also conducted to determine which major existing frameworks use these tools and techniques in the development of smart chatbots. Thus, Shortly after this systematic review and market research, the Rasa and Botkit frameworks were selected, as together they met all pre-established requirements. Finally, in this project the intelligent real estate agent LIA was developed.