This report discusses recommendations made to the different stakeholders involved in the adoption of Conversational Agents to set up learning discussions. These recommendations derive from the creation of agent and learning analytics technology, its integration with three different learning platforms and its evaluation along many pilots, which have been carried out in the project. Educators and instructional designers are advised to think about the design of CA-mediated debates at course level (their role in the course), task level (how they can foster productive conversation) and at the interaction level (how their interventions promote the desired skills). Technology providers should aim at a seamless integration of the agent chat with the learning platforms, keeping the educator in charge of defining the agent behavior, as well as connecting learning analytics from both technological platforms. Institutions should consider these activities not only in large scale courses. Finally, policy makers are recommended to promote the use of CA technology in formal education, especially in relation to the development of critical thinking and reasoning skills. They are also encouraged to look at the ethical implications of using human sensitive data, such as conversation, to intervene in the learning process and analyze it.
An exciting opportunity to learn about and participate in an international community devoted to the use of Conversational Agents for Learning. Conversational Agents (CAs) are an innovative technology that uses Artificial Intelligence techniques to provide more personalized and fruitful interactions for learning.