ABSTRACT
Advances in intelligent computing during the last decade have changed the world as we know it; possibly forever. Intelligent algorithms are constantly analyzing Internet traffic, diagnosing illnesses together with medical professionals, trading stocks, driving cars and suggesting what a person should watch on Netflix. Even in the field of education, intelligent algorithms and machines suggest suitable tasks and courses for students, analyze teaching, generate exams and even are able to predict the grade the student will gain after a course. Machine intelligence (or artificial intelligence) is taught throughout the universities as a part of computing degrees, but lately, machine intelligence has been taught in school levels below the universities (even in the kindergarten!). This, of course, creates a demand that students and teachers become literate in machine intelligence to satisfy the constraints and possibilities in modern digitalized world.
This thesis studies how machine intelligence can be taught efficiently for – what we call – data science novices. We show how usually complex and opaque intelligent algorithms and models can be taught efficiently through practice and by turning the opaqueness into transparency. Furthermore, this thesis studies how data science novices in the field of education can use machine intelligence and learn together with a machine through collaboration. We show how these novices can extend their perception, their attitudes and understanding by augmenting themselves and by collaborating with an intelligent algorithm in a way suitable for complete beginners. We call this ’augmented intelligence method’.
The implications of this thesis extend the current machine intelligence curriculum by providing easy approach for the beginners. Without knowledge in complex mathematics and statistics, the beginners can learn the concepts of machine intelligence and in the meantime, they can harvest the strength of one of the greatest inventions in the modern world. Data mining can be carried by the novices as well; the novices learn together with an intelligent algorithm; together they are able to build recommendations, discover knowledge in data, predict stock markets, predict grades or even detect fake news. The novices learn together with the machine, not by guidance of the machine. The perceptions of the machine and the novice support and correct each other. Through the augmented intelligence method, novices can use data mining efficiently with great depth and accuracy. Of course, the implications are not restricted to education, but the findings can be generalized in other fields as well.
The doctoral dissertation of MSc Tapani Toivonen, entitled Open machine intelligence in education will be examined at the Faculty of Science and Forestry on the 16th of October online. The opponent in the public examination will be Associate Professor Ryan S. Baker, University of Pennsylvania, USA, and the custos will be Senior Researcher Ilkka Jormanainen, University of Eastern Finland. The public examination will be held in English.
