| Summary: | Information about a student’s level or state ofknowledge is a key aspect for efficient, personalized learningactivities. E-learning systems gain such information in twoways: directly by examining users’ self-assessment andadministering predefined tests and indirectly by makinginferences on observed user behaviors. However, most of thecurrent solution approaches either demand excessivemanpower or lack required reliability. To overcome theseproblems, we have developed the e-Examiner, an assessmenttool that supports the assessment process by creatingautomatically test items, assessing students’ answers andproviding feedback. In this paper, we firstly give anoverview about a variety of computer-assisted andcomputer-based assessment systems and methods thatsupport formative assessment activities. Secondly, weintroduce the overall concept and architecture of the e-Examiner. Thirdly, we outline implementation details andevaluation results of our prototype implementation. Oursolution approach is based on the set of statistical similaritymeasures defined by the ROUGE toolset for automaticsummary evaluation.This paper is an extended version of the IMCL 2007 paper.
|
|---|
