In Washington DC, George Mason University convened a workshop on Big Data in Education, sponsored by the National Science Foundation,the Academy of Finland, and the Finnish Embassy. It brought together an eclectic group from education, computing and several who shared their experiences of how massive data is transforming their fields.
The challenge laid down as as follows:
How might we, from scratch, design digital platforms to model multiple data streams from multiple sources in a generalized ecosystem of learning to make predictions about learning based on changes to instruction?
We envision MOORs as digital terrains traversed by learners across formal and informal education (e.g., schooling, museums, the internet), and across the lifespan. For example, we will have experts on "experience API" (adlnet.gov) on tracking learning across multiple digital platforms.
The architecture could be extended to cover the use of data from performance assessments where the data originates from sensor nets worn by, and surrounding, the learner. We also want to explore data from biometric sources (e.g., "quantified self" data).
A few of us were invited to share our work — my slides…
Learning Analytics : quantifying deeper learning dispositions? from Simon Buckingham Shum
