Advancements in Intelligent Support for Collaborative Learning
Seiji Isotani - http://isotani.caed-lab.com/CAED - Applied Computing in Education Laboratory/Laboratorio de Computação Aplicada à Educação e Tecnologia Social
The approach comprehends 5 steps:
1) Take a real problem: That of Collaborative Learning
2) Organize the Knowledge
3) Build an Ontology
The work proposes that ontologies are used to formalize Collaborative Learning, considering different aspsects, for instance:
- Learning Strategy (Pedagogical theories): learning goals, learning roles etc.
- Domain Content Modeling
- etc.
He reports on lots of difficulties to convey the ontology to teachers, so they end up not using their method.
4) Hide the Ontology
However, they start proposing different models, which are simpler to understand. For example, the Learner's Growth Model (LGM), which is a graph that shows the stages of development, through empty and full squares.
The LGM maps cognitive theories to learning theory. Sooo interesting!!!
To take the student from an empty state (no knowledge) to a full one (good knowledge), you follow paths in the graph moving to states where learners learn new skills XOR new knowledge.
5) Apply the model and the ontology to solve the problem taken in 1)
LGM is a way of "hiding the ontology" so that people can understand the model and, at the same time, on the background, the ontology is still used for the benefit of the method.
Another interesting result is
They made a lot of in vivo that proves that the tool actually work in practice, leading learners who had poor academic results to better ones.
They created a startup named MeuTutor and today, they have 50.000 students in Brazil using their system.
Book in Portuguese:
It explains how to open the educational data for the community
In this direction, there is also this initiative from another group: http://meututor.com.br/blog/
Nenhum comentário:
Postar um comentário