Current students


GIUDICI MATHYASCycle: XXXVII

Section: Computer Science and Engineering
Advisor: GARZOTTO FRANCA
Tutor: CAPPIELLO CINZIA

Major Research topic:
Use of Persuasive Technology in Conversational Agents for Ecological Transition in Society

Abstract:
It is now undeniable that climate change is taking place. The Intergovernmental Panel on Climate Change argues (with about 95% certainty) that climate change is anthropogenic. The urgent need for a green turnaround of all nations on the planet has been highlighted, especially in all COVID-19 post-pandemic recovery plans. More and more people need to realize that a collective effort is necessary, also considering climate change's impact on future generations.
In Human-Computer Interaction, it is growing, and even is necessary, the interest in environmental sustainability. Several challenges must be addressed, and numerous issues must be well-considered and exploited. Nowadays, we have digital systems to prevent food waste, mobile applications to encourage sustainable mobility, and provide feedback on domestic electric power usage. Persuasive Technology using Conversational Agents is presented using visual interfaces to give feedback, e.g., about energy consumption. Effective persuasion, also in human-human interaction, is usually achieved using personalized messages and strategies, according to the subject to convince. However, thanks to advanced machine learning and data science techniques, persuasive technology could be achieved using algorithms that customize messages according to the user. Yet to be explored is the combination of IoT-enabled ambient technology and Conversational Agents (based on both voice-only and multimodal interactions) to address sustainability persuasion purposes beyond mere feedback that warns of inappropriate behaviors.
The research aims at developing a novel category of Conversational Agents, called Persuasive Conversational Agents (PCAs), to persuade people to behave more sustainably. Thanks to the integration of PCAs with the Internet of Things (which can capture a massive amount of data), PCAs can analyze environmental data. They can provide more targeted and contextualized feedback and implement more efficient persuasive techniques (e.g., social comparison, goal setting) that influence people's attitudes and behavior. The research will be characterized by: i) the definition of conceptual models for Persuasive Conversational Agents and their interaction paradigms; ii) modeling and developing a platform that integrates and extends existing conversational technology tools and IoT solutions exploiting the different conversational flows according to different persuasion techniques; iii) the experimental evaluation of the framework's effectiveness in the field.
Moreover, the research aims to have an impact on both technological and social aspects. From a technological perspective, the aim is to create a new category of Conversational Technology called Persuasive Conversational Agents, able to sustain persuasive conversation using advanced persuasion and argumentation strategies based on the knowledge about the environment enabled the integration with IoT devices. From a social perspective, the Persuasive Conversational Agents developed and tested during the research will contribute to growing environmental awareness, persuading people towards a green habit change, and promoting sustainability-led virtuous behaviors.