JISC CETIS Conference 2007 - Parallel session: Semantic structures for teaching and learning

At the JISC CETIS Conference 2007 I attended the parallel session Semantic structures for teaching and learning. The session prompted me to give some thought to the application of semantic technologies in education.

Referencing tools
One tool that came to mind during the semantic structures workshop was RefGrab-It (there did not seem to be an appropriate time to mention it on the day). For those who have not come across it RefGrab-It is a bookmarklet for Fireworks, IE or Netscape and it does exactly what it says – grabs basic bibliographic information from a web page (download it from: http://www.refworks.com/Refworks/BookMarklet.asp). The information captured includes page title, URL, and date accessed. It was developed for use with RefWorks, a web-based bibliography manager. I am not sure exactly how it works but it occurs to me that a tool like RefGrab-It could be much more effective if the data available to it were described more accurately in the first instance, or if gathered all the data needed for citation purposes. The results you get from using it do vary quite considerably. Perhaps the sometimes sketchy data retrieved from some pages could be checked against a richer data source (publisher database for example). This is unlikely to happen without financial exchange. I lie! According to RefWorks, “If an ISBN number, PubMed ID or DOI exist on the web page, RefGrab-It will automatically search various web resources to get supplemental information that may be of interest to you. RefWorks may also locate RSS feeds related to that web page.” However, it still does not provide full reference data for a web page even if that web page is an HTML version of a published paper. For example, author and journal details are not included unless additional information is embedded in the page. So, RefGrab-It is a potentially useful tool that falls somewhat short of expectations, but perhaps only because of the information available to it.

For a look at RefGrab-It and the additional info can capture (DOI, ISBN, RSS) try it on the following blog posting - http://www.scholarsportal.info/spot-news/?p=123. I have included a couple of screen grabs to illustrate.


Figure 1: RefGrab-It's basic bibliographic data


Figure 2: RefGrab-It's more advanced bibliographic data

Potential benefits to an improved RefGrab-It style tool
If we accept that the majority of learners (undergraduates / masters / PhD students, are researchers, academics, practitioners, etc) are using web searches (Google) to locate information (published papers, conference papers, etc) then being able to grab reference data directly form a web page to export into bibliographic software for later creation of a reference list will reduce the effort in the manual creation of such lists (and improve accuracy?). I guess you could argue that if learners are not using the traditional tools for locating papers and other texts then they are unlikely to be using bibliographic software like RefWorks. However, they can easily copy and paste from RefGrab-It into their own document. For teachers it could be used to illustrate best practice in gathering references: in information literacy training, citation, plagiarism, and so on.

This is probably more of a metadata issue but I suppose my question is can semantic technologies do a better job of getting this type of information from webpages and presenting it to learners?

Mapping tools
In his presentation Using Ontologies in eLearning, Michael Gardner introduced us to AUTODISCOVER. This tool can be used to make, and share, a 3D map of the documents on your computer. It struck me that the visual meta search engine Kartoo (http://www.kartoo.com/) does a similar job of bringing together your online documents. It also makes links / connections between your works and those of others (via the individuals you collaborate with). It would be interesting to see how AUTODISCOVER might bring computer and online collections of items and people together.