Unknown macro: {center}

Big Data Home

Skip to end of metadata
Go to start of metadata

The attached document is a draft proposal for a project to be proposed to the High Level Group for Modernisation of Statistical Production and Services (HLG) in October 2013, with a view to undertaking the project during 2014.

It was prepared by a Task Team of experts from national and international statistical organisations around the world.

Comments from all interested parties are very welcome.  Please use this discussion thread, or if you prefer to comment in private, send a message to steven.vale@unece.org with copy to fiona.willis-nunez@unece.org  




  • No labels

1 Comment

  1. On the whole, I think it is a good proposal. However, in workpackage1I would like to see more research into the different nature of most big data being transactional versus state descriptive. One of the consequences is that you can think of big data streams as signals and then the theoretical body of signal processing and signal recovery from noisy sources comes into play. It is important to realise that several higher levels of abstraction are present in those data of which recovery will not be trivial but,I think, worth our while.

    Another thing is correlation. We are used to looking for the correlations we can imagine. Big data has the potential to show existing correlations be it in or outside our imagination. This will enrich the statistical process. We cannot assume that correlations in the data that do not coincide with our own reasoning is faulty, we need to find ways to verify and use. Maybe some research can be beneficial.