"For addressing the first objective, we developed a semantic profiling and recommender system for scientific publications; implemented and tested the “Learning-how-to-search” and “Curriculum reflection” widgets of the Adaptive Training Support system, to guide users on how to efficiently use the platform and familiarize them with its technologies (crawling of social media, websites and videos; incremental schema-level indexing for distributed Linked Data retrieval; duplicate detection and removal; author name disambiguation; entity extraction and linking; detection and forecast of trending topics; automated semantic document annotation; video processing; user logging mechanism; and various visualisation technologies). Also, we developed the use case curricula for auditors and junior researchers, and produced learning materials and educational sequences. Furthermore, we implemented and evaluated a MOOC (Massive Open Online Course) as a practical implementation of the use case 2 curriculum, as well as new educational sequences.
For meeting the second objective, the consortium increased the amount of data resources available in the platform and developed a rich set of dedicated technology components (covering data acquisition, data processing, user logging and data visualisation), as well as a common data model for supporting the exchange between them; they also worked on the integration of these components in the MOVING platform. Regarding data acquisition, we improved the crawlers and extended the FLuID model. Regarding data processing, we extended our technique for duplicate detection, introduced two different techniques for entity extraction and linking, and developed a trend detection and forecasting method. About video processing, we extended the transcript-based lecture video fragmentation method, and developed a new machine learning technique for dimensionality reduction and concept annotation. Regarding the user logging method we improved the WevQuery tool, and evaluated the pattern mining features of WevQuery through a user study. Finally, regarding data visualisation, we improved the Concept Graphs, and added the uRank, the Tag Cloud and Top Properties user interfaces.
Concerning the third objective, the work done focused on the definition of the overall platform architecture, the separation of the test and production server, and the integration of all project functionalities and learning materials in the platform. More specifically, the most important developments were the complete redesign and adaptation of the project environment for community building, the integration of the MOOC and its functionality, the integration of the Recommender System widget, the Adaptive Training Support widgets, the user interaction tracking system, and all new and updated data acquisition and processing components, the development and integration of large-scale data collection techniques, and the design and implementation of the platform’s UI, including its responsive design.
Finally, addressing the fourth objective involved organizing and participating in a multitude of dissemination and networking activities, such as the presentation of MOVING results in more than 80 industry- and scientific-oriented events (conferences, workshops, exhibitions, summer schools, invited talks, demonstrations, open door days, user days, meetings and presentations). Also, the implementation of the MOVING MOOC represents a community-building effort on the MOVING platform. The exploitation strategy of the project focused on analysing the market environment and on guiding the research and development towards addressing the ""real-life"" needs of public administrators and researchers. This involved, for instance, using the business model canvas (BMC) methodology for capturing the exploitation parameters of the platform and all other IP assets generated in MOVING.
"
