Learning Task-Related Strategies from User Data through Clustering

Cocea, Mihaela; Magoulas, George D.

doi:10.1109/icalt.2012.92

Cited by 1 publication

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some strategies are used more frequently, and these can be introduced in the system by teachers. To ensure other relevant strategies are added to the knowledge base of valid solutions, we developed an approach which is described in [16], [18] and extended in [17].…”

Section: Learning With Expressermentioning

confidence: 99%

“…In our context, knowledge refers to approaches to individual tasks rather than concepts of the domain. over time were discussed in [16], [17], [18], and details on the overall methodology can be found in [19]. We argue that the proposed approach is particularly suitable for exploratory tasks and can offer a good balance between modelling under uncertainty and dealing with a wide solution space.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Design and Evaluation of a Case-Based System for Modelling Exploratory Learning Behavior of Math Generalization

Cocea

Magoulas

2017

IEEE Trans. Learning Technol.

Self Cite

View full text Add to dashboard Cite

Exploratory learning environments (ELEs) promote a view of learning that encourages students to construct and/or explore models and observe the effects of modifying their parameters. The freedom given to learners in this exploration context leads to a variety of learner approaches for constructing models and makes modelling of learner behaviour a challenging task. To address this issue, we propose a learner modelling mechanism for monitoring learners' actions when constructing/exploring models by modelling sequences of actions reflecting different strategies in solving a task. This is based on a modified version of case-based reasoning for problems with multiple solutions. In our formulation, approaches to explore the task are represented as sequences of simple cases linked by temporal and dependency relations, which are mapped to the learners' behaviour in the system by means of appropriate similarity metrics. This paper presents the development and validation of the modelling mechanism. The model was validated in the context of an ELE for mathematical generalisation using data from classroom sessions and pedagogically-driven learning scenarios.

show abstract

Section: Learning With Expressermentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%