Deep learning and sentiment analysis for human-robot interaction

n this paper we present an ongoing work showing to what extent semantic technologies, deep learning and natural language processing can be applied within the field of Human-Robot Interaction. The project has been developed for Zora, a completely programmable and autonomous humanoid robot, and it aims at allowing Zora to interact with humans using natural language. The robot is capable of talking to the user and understanding sentiments by leveraging our external services, such as a Sentiment Analysis engine and a Generative Conversational Agent, which is responsible for generating Zora’s answers to open-dialog natural language utterances.