A Mediabase for Technology Enhanced Learning in Europe
Abstract
There is growing interest in analytic techniques and tools for supporting scientific communities, e.g. through visual analytics tools. In the domain of Technology Enhanced Learning (TEL), one of the first endeavors in this regard started in 2006, when the PROLEARN project [1], a European TEL network of excellence, deployed a “Mediabase” for TEL in Europe. The objective was to overcome fragmentation and propel excellence in the field by providing stakeholder groups like scientists and policy makers with digital information obtained from TEL social media sources (e.g. mailing lists, blogs, RSS/Atom feeds) [3]. From the end-user perspective a key task was the development and provision of easy-to-use tools for extracting and presenting relevant information contained in the Mediabase, e.g. for cross-media social network analysis and self-observation [4]. In the scope of the currently running TEL-Map project [2], a support action funded by the European Commission, the Mediabase idea was adopted to empower TEL stakeholders in Europe and beyond to find relevant information and to obtain a rich overview of different types of actors and artifacts involved in the TEL domain. The main objective is to provide support forroadmapping activities in TEL, which required extending the available data sources and analytics tools.
Data Sources
The TEL Mediabase includes information from three main data sources:
- Blogosphere: TEL related blogs, including the blog entries, comments and analytical information like length, hyperlinks, words occurrences, and word bursts for entries.
- Projects: TEL projects funded by the European Commission, including metadata and information like description, start/end dates, cost, funding, and consortium members.
- Papers: Bibliographic information on TEL related conferences, journals, workshops, authors, and papers.
Figure 1 shows the component layers of the Mediabase. The web data sources are crawled periodically, and the data is processed and fed into a set of relational databases. The projects database is fed from publicly available information pages on the European Community’s websites (e.g. CORDIS [3]); the papers database is fed by the DBLP [4] data set enriched with abstracts and keywords crawled from the publisher pages (currently we support IEEE, ACM and Springer). The blogs database is fed by a blog crawler and a feed importer. The blogs to be included are specified by community members using web-based tools like the Feed Aggregator on the Learning Frontiers portal (http://learningfrontiers.eu) or the Mediabase Commander available at the PROLEARN Academy homepage (http://prolearn-academy.org).
Figure 1 - TEL Mediabase component layers
Currently the Mediabase indexes almost one thousand TEL related blogs with close to 400,000 blog entries and more than one million distinct hyperlinks. It includes the complete DBLP data set, as of March 2011 consisting of more than 800,000 authors, about 1.5 million papers, and more than 3,700 conferences, journals and workshops. The project data set includes 116 collaborative projects funded by the European Commission during TEL focused calls in FP6, FP7, eContentplus, and eTEN, involving more than 800 different organizations.
End-User Applications
To enable TEL stakeholders to access the data for analytic and roadmapping tasks, we deployed several web-based tools as shown in the bottom layer in Figure 1. In the Learning Frontiers portal, a TEL roadmapping portal hosted by the TEL-Map project, we offer the following tools:
- Projects Space: This portal section (see Figure 2) allows users to browse the project data set. It also presents a map overlay that shows the geographical distribution of project partners and an automatically generated list of predecessor and successor projects. These pieces of information are hyperlinked to provide a seamless browsing experience. A study investigating collaboration networks and project impact based on these features is presented in [2].
- Mediabase Dashboard [1]: A portal section similar to the iGoogle homepage. Users can embed widgets in their personalized dashboard. These special widgets provide real time visualizations of the data in the three databases. Users may also create custom visualizations by using the provided widget creator application.
- Feed Aggregator: Allows users to specify feeds and blogs to index for crawling.
Figure 2: Projects space on the Learning Frontiers portal
In addition we have developed and deployed web tools outside of the Learning Frontiers portal:
- Query Interface: provides the possibility to query the three databases using SQL in a web-based interface. The query results are visualized using different chart types like pie chart, line chart, timeline, table, graph, etc.
- AERCS [5]: enables users to access analytic, visual and comparative functions on the papers database. For instance it allows comparing the co-authorship and citation networks of different conferences, visualize author networks, identify key authors, and similar functions. A study analyzing the development patterns of TEL conferences using these features is presented in [5].
- Mediabase Commander: a Firefox plugin that allows users to browse the TEL blogs database and to add new blogs.
Conclusion
In this article we have presented a Mediabase for TEL, which was built and deployed to facilitate mapping and roadmapping activities in the TEL scientific community. This includes observing key indicators in relevant data sources as well as to contribute to and browse data sources from web-based applications within and beyond their community web portals.
Acknowledgments
This work was supported by the European Commission through the TEL-Map support action (FP7-257822).
References
[1] Derntl, M., Erdtmann, S., Klamma, R.: A Widget-Based Dashboard for Visual Analytics on Scientific Communities. Proceedings of I-KNOW 2012. ACM (in press)
[2] Derntl, M., Klamma, R.: The European TEL Projects Community from a Social Network Perspective. Proceedings of EC-TEL 2012. Springer Verlag (in press)
[3] Klamma, R., Spaniol, M., Cao, Y., Jarke, M.: Pattern-Based Cross Media Social Network Analysis for Technology Enhanced Learning in Europe. Proceedings of EC-TEL 2006, pp. 242-256. Springer (2006)
[4] Petrushyna, Z., Klamma, R.: No Guru, No Method, No Teacher: Self-classification and Self-modelling of E-Learning Communities. Proceedings of EC-TEL 2008, pp. 354-365. Springer (2008)
[5] Pham, M.C., Derntl, M., Klamma, R.: Development Patterns of Scientific Communities in Technology Enhanced Learning. Educational Technology & Society (in press).
Michael Derntl
RWTH Aachen University
Germany
Ralf Klamma
RWTH Aachen University
Germany
0 comments:
Post a Comment