JOURNAL OF COMPUTERS (JCP)
ISSN : 1796-203X
Volume : 2    Issue : 3    Date : May 2007

Hybrid System for Generating Learning Object Metadata
Olivier Motelet, Nelson A. Baloian, and José A. Pino
Page(s): 34-42
Full Text:
PDF (1,653 KB)


Abstract
Generating LOM for learning material is a complex and tedious task to be manually completed.
Therefore, the current trend is to automate this process. However, there are several important
issues restricting this approach. First, all the necessary information for generating metadata values
is not always available. Second, some LOM attributes are difficult to generate in a rational and
objective manner. In order to cope with these issues, this article suggests considering human
intervention in the LOM generation process supported by automatic processes: a hybrid system in
which both human and computer collaborate for instantiating LOM attributes is described. In this
work, a tool implementing this proposal is implemented as a feature of LessonMapper2, which is a
graphical editor designed to author lessons as graphs of learning resources.

Index Terms
learning-object metadata, user-interface for metadata generation, automatic metadata generation,
lesson graph.