Genius is eternal patience.
Concept, UI design
Martinus is a personal assistant chatbot based on Artificial Intelligence (AI). Martinus is controlled in conversational form via a mobile application similar to WhatsApp. The application consists of two parts: learning and assistance.
This project was created for Tilburg University. The objective was to create a concept for the Smart Campus of Tilburg University. Our goal was to design something innovative in the category ‘smart learning’.
“Quartz's new app puts the entire global economy in your pocket. It's an ongoing conversation about the news, sort of like texting. We'll send you messages, photos, GIFs, and links, and you can decide when you're interested in reading more. Each session lasts just a few minutes, so it's perfect for the train, elevator, grocery store line, or wherever you have a spare moment.” Seward, Z. (2016, February 11). It’s here: Quartz’s first news app for iPhone. Retrieved from qz.com.
Our concept is based on the ‘interaction through conversation’ concept of the Quartz news app. Instead of traditionally pushing buttons, interaction takes place in a conversational way like messaging someone via Whatsapp. When researching the subject, we encountered a news article by TechCrunch stating: “Why Quartz’s news app is so much bigger than news.” Popomaronis, T. (2016, Mar 22). Retrieved from techcrunch.com. So we asked ourselves: Why not create something similar for educational purposes? As an answer to this question, concept Martinus was born.
Martinus helps students in their studying by examining relevant course material in preparation for a test or exam. Examinations come in multiple choice form, which is a closed input method. The student can be examined at spontaneous moments via pop-up notifications or whenever he or she has some spare time. Just like in the Quartz app, examination sessions are brief. Learning material can be gathered via Martinus Management, an external web application where students and tutors can collectively create classes, subjects and questions. Content is accessible for anyone inside the application and it is possible to suggest modification to content owners. This enhances the quality and reliability of the overall content.
Martinus assists students by helping them with daily repetitive tasks. Assistance works with an open input method, which lets the student ask questions in their own way. For instance, a student might ask Martinus about his or her study schedule, building locations, time management and relevant news about the university. Another possible feature is event based interaction: whenever a student gets a bad grade, Martinus could ask why this happened and help the student improve his or her knowledge by examining extra study material on that particular subject. In this way, learning and assistance are connected and synchronised.
The application design is comparable to that of many traditional messaging applications. All interactions - except for learning, settings and Martinus Management - are controlled in a conversational form. At the bottom left corner, the user finds a hint feature where all possible interactions and context are listed. This will help the user enlarge his or her knowledge on how to broadly communicate with the application. A connection can be created which enables the user to log into Martinus using his or her own university account.
By interacting with different students, the AI will learn more over time. During interactions, data is generated which can be used to improve the individual student learning and assistance paths. Based on interaction patterns, predictions can be made on future study results, student motivation and progress. This data can be sent to tutors and supervisors later on.