Specifications, models and standards for personalisation features and inquiry learning apps

Personalisation of the Go-Lab front end solutions is critical to improve usability, findability and user experience. In this deliverable, we analyse the requirements for personalisation through use cases that detail how teachers can be assisted during their activities by these functionalities. Afterwards, we provide comprehensive surveys on existing standards and specifications that enable personalisation. We first discuss the data needed to enable personalisation. Labs, inquiry learning spaces (ILS), apps and learning resources will include rich metadata on top of their content that can be used for effective filtering and recommendation. For apps, Go-Lab will follow the OpenSocial metadata specification and the ROLE Ontology. For resources, ILS and labs, metadata specifications are still under dicussion. Then, we present specifications for personalisation in Go-Lab. More specifically, personalisation in Go-Lab will be centered around internationalisation and recommendation. For internationalisation, Go-Lab will support the personalisation of languages at ILS creation time. For recommendation we propose to use a hybrid recommender system using collaborative-filtering and a multi-relational graph used in Graasp. We also propose to investigate federated and time-sensitive recommender systems. Finally, we present specifications for inquiry-learning apps, which will be based on OpenSocial standards for communication and data storage & retrieval. The apps will target both desktop and mobile devices. Discussions on software libraries that enable the use of apps on both desktop and mobile clients are still on-going. This deliverable prepares the specifications, models, and standards for personalisation which are the foundation for the specification of the Go-Lab Portal (D5.2).