Metrics Based on Attention Metadata for Learning Resource and Assessment Repository

Considerable effort has been put in place in recent years to build repositories of learning resources and assessment activities. Thus, raising the demand to collect and curate meaningful metadata related to their use. This is order, to improve user experience leveraging on educational data mining techniques. This work will propose firstly a set of data that can be collected and a set of similarity metrics useful for query recommendations on based on their content. Secondly, the set of results user recommendations either based on the queries they perform on the system, or on the basis of the activities they perform on the repository. Thirdly, learning resource recommendation based on other users' interests and results of similar queries, pattern of usage of resources and feedback given by similar users. An activity matrix collecting data in three dimensions: resources, users and queries is proposed; leaving implementation details and experience report for further studies.