Final version of the platform, including additional technology required for dealing with OER, cross-lingual, cross-site and cross-media data, as well as completed integration with components developed or under development in WP3 and WP4.
Extension of recommendation engine to support cross-lingual recommendations for multilingual users.
Final version of the user modelling prototype specifically focused on OER and taking advantage of rich meta-data and behavioural data available through OER social network.
Final version of the visual interface, including exploratory interface for analysing the platform and user behaviour.
The integration of further and richer information from the social network, and of topic maps will allow to build more complex alignments between learning traces and educational resources. Multi-lingual dimension is taken care with: courseware of different modes and languages can be aligned together.
Building the alpha version of the system. This version will rely essentially on language modelling for the conception of the different maps. Building of evaluation tools and methods to monitor the progress to be made with the different choices.
Final version of the three products and services including corresponding APIs.
Extension of recommendation engine to better handle cross-site OER and users.
First version of the user modelling prototype, mostly based on pre-existing news-recommendation algorithms.
Final version of the recommendation engine specifically focused on OER and taking advantage of rich meta-data and behavioural data available through OER social network
Extension of recommendation engine to support cross-site recommendations for multilingual users.
A finer analysis of contextual information is proposed allowing the alignment to correspond to intent: topic maps allow to recover the type of question a learner may want to make at any given time of his learning experience. The system is fully operational and running.
Construction of advanced content representation models and publication of the models at a conference/journal.
Final version of the cross-lingual component, focusing on translation of OER such as subtitles of videos.
This deliverable will report on the on the final analytics infrastructure and the second set of tools.
Report on the privacy and copyright issues relevant for the project.
This deliverable will report on the devised piloting scenarios and first piloting on the first infrastructure and first set of tools.
Building evaluation methodologies for measuring quality of content representation models and quality assurance models, publication of the models at a conference, and making labels obtained for measuring OER quality publicly available through the release of a dataset
This deliverable will be based upon a study of which “real world” activities (meetups, focus groups, user analytics), the social media channels (mailing lists, wikis, blogs) to be strategically targeted by the project, and will plan a series of X5gon organized events. A plan will be put in place for feedback solicitation and continued engagement. In particular we will propose how impact related to those channels will be monitored throughout the project’s timeline.
This deliverable will report on the complete 5Xsite participation in online, external learning and dissemination events, social media engagement, dissemination of the marketing materials, and the feedback received. A longevity plan will outline successful events and their potential adoption by new projects and collaborations.
Construction of quality assurance models and publications of the models at a conference/journal.
Final evaluation of the platform and connected components
The Project Handbook aims at supporting X5gon partners to manage all the administrative and organizational aspects of the project. The document provides introductions to the overall administration and reporting procedures, referring to the official EC documentation and to the contractual documents where details can be found. Other aspects treated in the document are project management and coordination procedures.
Report on existing technology available within consortium, as well as the core requirements for the platform with respect to the social network (WP1) and plugin components (WP3 and WP4).
This deliverable will report on the on final solution implemented at the extended set of user sites.
The report will include the findings from in-the-wild evaluation of the service performance and studies about the adaptive learning experience of both virtual and real world meeting groups. The results will feed back to WP1-4.
Construction of initial content representation models and publication of the models at a conference/journal.
This deliverable will report on our participation in online and external events, and learning and on the delivery of, and feedback from these events. The plan from D7.2 will be revised in light of this feedback.
Report analysing the quality of the quality assurance models built.
Through collaboration with both internal partner organisations and external potential customers, a set of value-add metrics will be prepared, and a report produced detailing the reasoning. For each iteration, M12, M24, M36, the progress of the project will be assessed against the criteria, and the feedback distributed to all participants.
This report summarises the key findings of the evaluation studies of the OER network and user interface designs. The findings and recommendations will inform the next design decisions and revisions for technical WPs.
Report analysing and comparing the quality of the content representation models built.
The third report focuses on the in-the-wild study about final products, installations, OER sites experiences, students’and teachers’ experience and incentives for the OER sites community to join the network.
First version of the cross-lingual component, focused on shallow language-independent representation based on canonical correlation analysis.
The deliverable will comprise two sections the first on Ethical data management will cover: * Report on the privacy policies that will be used by the project and the different OER sites. These should specifically include the use of user data for research purposes * Details of consent gathering for minors (forms and data management policy) * Ethical clearance from the relevant DPAs or existing ethics boards for the activities in different countries * Report on the legality of the final product under GDPR * description of the algorithms used for machine learning and their ethical implications * identify data controllers and processors The second section of the report will be the Data Management Plan as part of the Open Research Data Pilot and will cover: the handling of research data during and after the end of the project * what data will be collected, processed and/or generated * which methodology and standards will be applied * whether data will be shared/made open access and * how data will be curated and preserved (including after the end of the project).
Final update of the report.
Update of the year 1 report.
The website will be cross linked from/to other relevant EC, OER and community sites. The website will be maintained and updated for at least 5 years after the project has ended. Documentation manuals and screencasts shall be specifically identified as project deliverables and prominently published on the website.
Author(s): Stefan Kreitmayer (UCL), Yvonne Rogers (UCL), John Shawe Taylor (UCL)
Published in: BCS, The Chartered Institute for IT (ACM Proceedings) 32nd Human Computer Interaction Conference, 07/05/2018, 2018
Publisher: The Chartered Institute for IT