Enhancing just-in-time e-learning through machine learning on desktop context sensors

  • Authors:
  • Robert Lokaiczyk;Andreas Faatz;Arne Beckhaus;Manuel Goertz

  • Affiliations:
  • SAP Research CEC Darmstadt, Darmstadt, Germany;SAP Research CEC Darmstadt, Darmstadt, Germany;SAP Research CEC Darmstadt, Darmstadt, Germany;SAP Research CEC Darmstadt, Darmstadt, Germany

  • Venue:
  • CONTEXT'07 Proceedings of the 6th international and interdisciplinary conference on Modeling and using context
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

The objective of novel e-learning strategies is to educate the learner during his actual work process. We focus on this new approach of in-place and in-time e-learning, which offers learning resources right in time the user is in need for it. A crucial factor for those modern task-oriented e-learning software is the user's context. To deliver learning resources to the user, which are both suitable and helpful with regards to the user's current work situation and his competencies, the application always has to consider the learner's actual work task, his environment, and history. In this paper, we present an architecture for the work task prediction, evaluate different machine learning algorithms in depth by their accuracy for that purpose and discuss the integration in our e-learning environment. This validates the possible usage in real-world business scenarios.