< Back to previous page
Project
Individual feedback in electronic study environments.
One of the essential aspects of learning is feedback (Kulhavy and Wager 1993, Mory 2003, Passier and Jeuring 2004, 2006). State-of-the-art e-learning systems discover language mistakes based on string matching techniques and are able to show the user the correction for their mistakes. They lack however the possibility to explain this correction and give valuable feedback. Mory (2003) already stated that there is a frequent lack of feedback in electronic learning environments. The little feedback presented in current systems is mostly hard-coded, not very valuable and almost always limited to multiple choice questions (Passier and Jeuring, 2004). There is however no denying to the fact that there is a great need of more and better feedback:
An eLearning system that produces semantically rich feedback is very desirable, because feedback is crucial in effective learning, feedback is sparse in most eLearning systems, and the number of eLearning systems and eCourses is growing rapidly. (Passier and Jeuring, 2004)
In this dissertation, we develop an automatic and adaptive machine learning algorithm capable of generating feedback for a variety of language exercises. The algorithm is initially trained with a large amount of data, but its main advantage lies in the fact that it improves itself based on newly inserted feedback. The quality and level of detail of our training and test data is constantly improved by our expert users (teachers), which are very willing to label our data.
An eLearning system that produces semantically rich feedback is very desirable, because feedback is crucial in effective learning, feedback is sparse in most eLearning systems, and the number of eLearning systems and eCourses is growing rapidly. (Passier and Jeuring, 2004)
In this dissertation, we develop an automatic and adaptive machine learning algorithm capable of generating feedback for a variety of language exercises. The algorithm is initially trained with a large amount of data, but its main advantage lies in the fact that it improves itself based on newly inserted feedback. The quality and level of detail of our training and test data is constantly improved by our expert users (teachers), which are very willing to label our data.
Date:1 Oct 2009 → 1 Oct 2013
Keywords:Feedback, Education
Disciplines:Education curriculum, Education systems, General pedagogical and educational sciences, Specialist studies in education, Other pedagogical and educational sciences
Project type:PhD project