WP1 – Data Capture


The main goal of WP1 is to address the challenge of data capture in online social environments for the purpose of learning analytics. Besides collecting basic social networking data, the issues here relate to, first, understanding/engineering the features in these data that can be used to identify and recognise learning activities across cultures, languages and learning practices, and second, to classify/re-purpose these activity data so that they can be meaningfully exploited to visually analyse the behaviour of informal learners (WP3), as well as to populate the cognitive models of informal learning developed in WP4. Accordingly, we will employ a layered approach to data capture, where data is first extracted raw from the social platforms and refined until it can be robustly used as evidence of informal learning within the learning analytics process.

Generally, WP1 will provide the base data management platform for the whole project, creating an interface between the raw multi-lingual data from social environments and the analysis/modelling of these data in other work packages

WP2 – Data processing and enrichment for analytics


WP2 will process and align data obtained from WP1 to make them available as part of a unified knowledge graph which enables exploration and analytics in WP3 and WP4. A range of automated and semi-automated techniques – schema mapping, entity linking, data fusion/integration, clustering – will be developed and specifically adopted to AFEL-related data and entity types, to first interlink and align all data into a coherent graph, which can be exploited and analysed, for instance as part of WP3. In particular, WP2 aims at detecting, emphasizing and enriching concepts (activities, events, entities) of relevance to informal learning, and, in particular, the ones identified in the cognitive learning models of WP4. Expected outcomes include (i) well-connected, large-scale graph/knowledge base of AFEL-related data, described according to linked data principles which can be exploited for further learning analytics and analysis activities, (ii) a set of novel data enrichment, alignment and linking methods, tailored in particular to data types relevant to AFEL, (iii) data processing and enrichment methods aimed at detecting and emphasizing learning-related notions and concepts in heterogeneous datasets, which will be informed by the cognitive (informal) learning models and pattern described in WP4.


While the lack of well-described, large-scale data is one of the crucial challenges for learning analytics research and practice, the outcomes of WP2 will form a substantial contribution to foster large-scale learning analytics especially in the open environments of online social platforms.

WP3 – Visual Analytics


The overall goal of this work package is to generate and evaluate tools for the visual analysis of informal learning networks based on cognitive models of learning. We aim to show relationships between different types of entities (e.g. learners, resources, and topics) in the knowledge graph constructed in WP2 and enrich this graph with indicators for informal learning in WP1. The graph visualization will aim at providing learning platform providers with the necessary insight to detect and mitigate blocks to learning. As a basis for the visualization, we will use findings on cognitive models gathered in WP4.
Expected outcomes include (i) a conceptual framework of visual analytics in informal learning, (ii) a set of tools for visual analytics of communities and community processes with regards to informal learning, and (iii) insights on the usefulness of visual analytics tools to foster informal learning.

WP4 – Cognitiv Models


This WP aims at taking up the rich data and the visual clues developed in and provided by WP1, WP2, and WP3. These data will be analysed and interpreted through the lenses of the co-evolution model of cognitive and social systems (Cress and Kimmerle, 2008). The analysis in WP4 will elaborate on those aspects which are characteristics of online
knowledge construction.

In a first step, it has to be understood how collaboration takes place between individuals and the collective as a whole. What triggers people’s willingness to contribute to the knowledgeconstruction processes? What role do socio-cognitive conflicts play in this context? What are incentives in varying contexts, such as resource-centric social networks such as Slideshare or Bibsonomy and task-centric social networks such as Crowdflower and Amazon MT.
If we want to understand how communities advance processes of knowledge construction, we need to describe and analyse how they organise themselves. How do the communities take up information from the environment and apply them for self-reproducing the communication? What rules do communities develop for ensuring its continuity?
If we want to understand how new ideas and new knowledge may develop, we need to analyse how knowledge-constructing communities handle complexity. How do they select those pieces of information that are relevant for them and filter out the rest? How do they reduce sources of irritation?
Finally, an examination of online social environments as knowledgeconstructing communities must include an analysis of processes that allow the individuals involved and the community as a whole to constitute meaning and assure a common understanding. How does the community make sense with regard to the topics discussed and how does this develop over time? How does the knowledge constructed by the community integrate different viewpoints?

WP5 – Use case and evaluation


WP5 aims at developing a use case to demonstrate the tools, techniques and models created in the project on real world, realistic social data, as well as to show their potential in terms of improvement of the support for informal learning (objective 3 of the project). The study will be mainly carried out within GNOSS-Didactalia, a social and semantic platform of linked networks with end-user applications oriented to the creation of social and informal learning spaces, as well as collaborative working environments. It will also be extended to other social platforms connected to Didactalia powered by GNOSS, as well as external ones as a way to generalise the findings from the project to other potential application environments. We will specially focus our attention here on the Jisc Learning Analytics architecture, LearnWeb and BibSonomy. Both LearnWeb and BibSonomy are hosted and developed by L3S Research Centre. While LearnWeb is a social learning platform similar to GNOSS, Bibsonomy ( is one of the top-three social bookmarking systems for collaboratively tagging web links and  publication metadata.It has been developed and is operated by L3S Research Center since 2006. Its more than six thousand active users have contributed and annotated more than 1.5 Million resources, among them metadata for almost 900000 scientific publications; it handles millions of search requests per day.

WP6 –  Dissemination and Exploitation


The goal of WP6 is to ensure the visibility of the outcomes of AFEL, both in terms of the dissemination of the results of research and development activities in the project, as well as in terms as their adoption through exploitation (commercial or otherwise).

Thus, WP6 will ensure the appropriate scientific, industrial and end-user dissemination of the results of the project. This includes organising the presence of the project at relevant scientific and industrial events, setting up and maintaining various channels for the project’s external communication, as well as organising user-groups to collect feedback and
reach out to potential user communities. WP6 will integrate the results of the project within the commercial offering of GNOSS. This corresponds to the direct exploitation of the tools, techniques and models created by AFEL as extensions of the GNOSS and other platforms (as developed in WP5). Finally, WP6 will identify and purse other exploitation channels. Based on the experience with GNOSS, the project will integrate the developed features for supporting informal learners in online social network as applications of other social platforms and investigate the economic potential of such applications.

WP7 – Project Management


This work package is dedicated to project management. The overarching goal is to ensure the timely delivery of the work, within budget, and to a high standard. To achieve this, the work package will define and maintain the work plans and monitor progress against milestones. It will coordinate communications both with the European Commission and within the consortium, define the quality assurance and risk management plans, and manage the progress of the project and the financial reporting tasks. The management structure and procedures for the project are detailed in section 3 of this proposal.