News Story

LTfLL project takes textmining technologies in e-learning further

Fridolin Wild, Wednesday 13 Apr 2011

Last week’s final review of the EC-funded ‘language technologies for lifelong learning’ (LTfLL) project found that technology is mature and ripe for the market.

Over the course of three years, the collaborative project LTfLL has developed a set of interoperable services that use natural language processing techniques and semantic web technology to support all aspects of individual and collaborative knowledge building: positioning, monitoring conceptual development from writings, social retrieval of learning resources, summary writing with automated feedback, and learning analytics for chat & fora.

KMi has been leading the work package on learning and language technology infrastructure, thereby focusing on widget-based personal learning environments (for the learning technology) and a data management & processing infrastructure that brings together latent semantic analysis with network analysis (on the language technology side).

In close collaboration with the Vienna University of Economics and Business, the Open University of the Netherlands, and the Polytechnical University of Bukarest, the work has contributed to the further development of the Apache incubator project Wookie and has resulted in several new releases of the lsa and related packages for the statistical language and environment R. Amongst others, this included a sparse implementation and accompanying web services that support large corpora up to millions of documents.

The developed infrastructure tools and the PLE widgets (and their underlying services) are released on SourceForge under open source licenses.