If there is one scenario that was prominent in driving the development of Linked Data at the Open University, it is the one related to the discovery of educational resources. Indeed, there is a basic assumption that providing structured, open and addressable descriptions of resources helps making these resources more visible. In fact, most of my early presentations of LUCERO (but, for some reasons, not the ones that are online) included a picture of somebody saying “I’ve just seen a very interesting BBC programme. What is there at the OU that can help me learn more about it?”. Two years later, we actually have a systems that does exactly that!

Indeed, with support from the Open University’s “Open Media Unit”, we built an application that can semantically analyse the textual content of online resources and match it agains semantically indexed Open University content (OpenLearn Units and Podcasts at the moment) . The result (implemented as a set of REST services, some Javascript and a bookmarklet) is, if I might say so myself, super cool. It’s called:


(and yes, we probably should have put more effort in choosing the name).

The whole thing is pretty much a combination of linked data and information retrieval technologies. The Open University resources are crawled through data.open.ac.uk, analysed using DBPedia Spotlight and indexed using Apache Lucene. A BBC programme page used as a starting point would pretty much go through the same process, using the RDF description of the programme from the BBC website, analysing the textual components and matching the results to indexed resources. Because we use DBpedia Spotlight, the resources are described (and indexed) based on DBpedia entities, which allows us to semantically characterise their overlap, based on the links between common entities. It also makes it possible for the user to customise the search process based on his/her own interests.