LAK12 Reflections – Week One

Share this post

Reading Time: 4 minutes

So, I’m having another attempt at attending a MOOC! I registered for the hugely popular Stanford AI Class last year but those attempts were scuppered by the predominantly video-driven course being delivered as Flash Video only, which made it impossible for me as my main learning device these days is an iPhone. Then last week I stumbled across Learning Analytics and Knowledge 2012, or LAK12 for short, an 8 week introduction to Learning Analytics led by George Siemens and hosted by the Society for Learning Analytics Research.

With my appetite for participating in a MOOC still strong, I hastily signed up as it had literally just started. Every week a number of tasks are set and and students are encouraged to post reflection on their blogs. So I’d like to welcome you to my LAK12 reflections!

Educational Data Mining

I downloaded their first resource (The State of Educational Data Mining in 2009) and read this on the train on the way up to the Learning Technologies 2012 conference, and was surprised to learn that Learning Analytics has been around for over 15 years in the form of Educational Data Mining (EDM). As someone who thought learning analytics was an emerging field, this came as a big surprise.

It seems that 1995-2005 are described as the ‘early years’ in EDM and a fair amount of academic research was done in this period.  A review of the research was done in 2009 by Romero and Ventura, who identified the most influential articles in EDM to date by the numbers of citations shown in Google Scholar. These provide a worthwhile reading list and it would be interesting to see how the field has progressed in the last few years with another Google Scholar investigation.

Research around EDM is growing fast. The 2008-09 EDM conference papers outnumbered the entire body of research published in the 10 year period to 2005.

Pittsburgh Science of Learning Center (PSLC) Datashop

It seems like the Pittsburgh Science of Learning Center (PSLC) Datashop is at the heart of the EDM research community. PSLC makes data from online learning environments freely available to researchers and accounts for 14% of all research in this area (a further 12% of research uses open VLEs like Moodle). This lowers the barriers to doing EDM research and makes research verifiable by others. The PSLC DataShop also provides a central repository for the learning science community to secure and store research data; and provides a set of web-based analysis and visualisation tools for researchers.

Researchers can access standard reports such as learning curves, as well as browse data using the interactive web application. To support other analyses, the DataShop can export data to a variety of formats that can then be used in statistical software and other analysis packages. Toward collecting data in a uniform format, PSLC developed a standard XML logging format and two logging libraries (one in Flash ActionScript, the other in Java) to write the XML. Data can also be imported using a similar tab-delimited format. After importing data or logging the data to the DataShop database, the DataShop web application can help you start exploratory data analysis with tools for common learning science analyses. You can also export data for further manipulation and analysis in other tools.

In short, it looks like a great resource and I hope to make use of it when trying out the tools we will be playing with later on in the LAK12 programme.

Gates Foundation to fund further EDM research?

The other article that stood out for me in this week’s reading was an Inside Higher Ed article: Technology and the Completion Agenda. This mentions the Moodle distribution, Joule, which features analytical tools as part of a “wrapper” around the core open-source Moodle learning platform. It also points out that the Gates Foundation’s announcement of $20 million in new grants for technologies that point toward student completion, for which learning analytics projects will be eligible. So maybe the Gates Foundation will fund research into this field going forward.

Mining social networks

Finally, an interesting 2010 article from The Economist showed how analytics are used in the wider field, specifically within telecomms and law enforcement using network analysis tools. Sites like Facebook and Twitter are analysed and merged with data and chatter from other sources to uncover “non-obvious relationships” which results in predictions about all sorts of things with alarming accuracy, from illegal raves to Hezbollah rocket attacks, as well as detailed insights into the social networks of criminals and other wanted characters, and apparently contributed to the capture of Saddam Hussein.

This of course is the tip of the analytics and data mining iceberg. Listening to the keynotes at Learning Technologies 2012 really hit home to me how important this area is. Ray Kurzweil, in an exciting and upbeat presentation, firstly described how the vast collection of data that the likes of Google and Wikipedia were building was an enabling factor in achieving the Technological Singularity, the moment whereby computer intelligence surpasses human intelligence, and he described how the people behind organisations like Microsoft and Google who are collecting all this data were obsessed by this idea to the point of investing in and building the Singularity University. The following morning, Jaron Lanier expounded on the flaws of such widespread and carefree contribution of personal data to these organisations, to whom millions of people contribute for some short term value but no long term gain other than making the Google and Facebook founders extremely wealthy. His message was to think before you contribute freely to these sites. Is it really for the greater good? Should we not be economically rewarded for our contributions?

Whether you take a utopian or dystopian view of these vast and rapidly growing pools of personal data, there’s no doubting that it’s data mining technologies and analytics tools that allow sense to be made and value derived from all of this data. It’s fascinating stuff and it’s great to be learning about how this all impacts on workplace learning and development through LAK12.

Share this post