- Posted By: Susan Dutton
I attended this symposium which aimed to provide researchers in Oxford with an overview of what Big Data is and how researchers can use tools that are being developed both within Oxford and further afield.
One speaker expressed the thought that we have always been dealing with Big Data. In the past we collected lots of data and carried out Data Mining to find interesting ideas that could then be tested in future studies. It wasn’t very successful as the data sets tended to be fairly small and the technology not very sophisticated. Now the data sets are much larger and the technology is rapidly changing. But it is essentially still data mining and any findings should always be tested in different data sets or new studies.
Dr Susanna-Assunta Sansone, Associate Director of the University of Oxford e-Research Centre, talked about FAIR data which is defined as Findable, Accessible, Interoperable and Reusable. Big Data is not just about the technology. The data needs to be stored carefully with sufficient details to allow future researchers to know what the data is and how it was created including the techniques that have been used to carry out the experiment and the processing that has occurred to create the final summary end-point. A tool kit is being developed to facilitate the process (see http://isa-tools.org/ for further details).
She is also working in collaboration with the Nature Publishing Group to enable the peer-reviewed publication of the description of the methodology behind this “meta-data” to complement the research article.
Although she was talking about basic science, her points were very pertinent to data sharing in clinical trials and how we should store the data for the future. Funding bodies now require that data collected in clinical research should be made available and some of the new techniques being developed may provide ways to make this easier.