We are proud to introduces the practical empirical chatbot system named SMARTA, which stands for “Student Motivation and Reflective Training AI-Assistants”. SMARTA is an ongoing project which uses the ChatGPT interface to seamlessly integrate chatbots into the student’s academic environment, specifically within their campus management system (CMS). Over 5000 students at a German university have been provided with this service since August 2023. They have access to three specialized chatbots: Alix, Robyn, and Dr. Melly.

  • ALIX (Affirmative Life Impact eXpert) actively engages students to motivate them through personalized interactions, aiming to enhance wellbeing and mental health.
  • ROBYN (Reflective Overview By Your Notions) assists students in critically reflecting on their academic and personal progress by fostering their curiosity in their specific field of study.
  • MELLY (Mentored Education with Logical Learning Yield) focuses on enhancing logical and critical thinking, and offers learning content based on general university study materials in an interactive way.

A unique feature of SMARTA is that the chatbots, through integration into the CMS, already know basic data about the student at the start of the conversation, such as age, gender, major, current study load, or even academic performance. The chatbots also have access to specific general information like university specifics or module handbook contents. They can initiate conversations with questions like, “Did the lecturer also cover the topic of advertising in today’s marketing lecture?”

Another special aspect of SMARTA is the intensive monitoring and recording of chatbot usage. While specific conversation contents aren’t recorded, data such as the length of the conversation in characters and time, as well as the sentiment during the dialogue, are tracked. This allows for analyses based on variables like sentiment change in relation to gender and field of study.

There are plans to enable students to save their previous dialogues in a keyword-based history, which can then be incorporated into new dialogues. For example, the chatbot could start a conversation with, “Did you find today’s marketing lecture boring again? Or did my motivation hint from the other day help?”