Analyzing navigation data to design adaptive navigation support in hypertext

摘要

Adaptive mechanisms in educational hypertext systems should scaffold learners to make the right choice for a particular task at hand. In this paper, the design of this type of support is discussed. The paper presents an empirical approach to designing adaptivity, in which a non-adaptive version of the system was used in a real classroom setting. Navigation patterns were then analyzed using the Pathfinder algorithm and the k-means clustering algorithm to group navigation patterns into clusters. The patterns in the clusters helped understand the kinds of paths that students take, which informed the design of conditions for adaptivity.