Raising up Annotations In Pedagogical Resources by Human-Computer Collaboration
Abstract
Huge repositories of pedagogical resources such as the French initiative of digital thematic libraries are now accessible by both students and professors. This high quantity of data makes their access difficult to the students, as they cannot get easily pertinent pedagogical resources that fit their needs. One way to make this access easier is to add annotations to these resources and to exploit these annotations to find pertinent answers. Of course, these annotations can be semantic. Semantic wikis are a new approach that automatically processes semantic annotations and that can be used to find the adequate resources given the requests of students. However, annotating semantically resources is not an easy task for human. Despite of their high potential, semantic wikis suffer from a lack of human provided semantic annotations, resulting in a loss of their efficiency. We propose a system (called HCA) that suggests automatically computed annotations to users in semantic wikis. In this paper users are students and professors. Users only have to validate, complete, modify, refuse or ignore these suggested annotations. Therefore, the annotation task becomes easier, and we assume that more users will provide annotations, leading to an improvement of the system and a facilitated access to pertinent pedagogical resources. The HCA system is based on collaborative filtering recommender systems, it does not exploit the content of the pages but the usage made on these pages by the students and the professors. The resulting semantic wikis contain several kinds of annotations, with different status: human, computer or human- computed provided annotations.
Domains
Artificial Intelligence [cs.AI]Origin | Files produced by the author(s) |
---|
Loading...