SES Big Data: Innovation in Analysis, Usability and Curation

The SES Big Data: Innovation in Analysis, Usability and Curation event was held on 18th September 2015.  Featuring talks from key researchers at SES institutions and industry, it offered academics the chance to showcase their research in the context of issues in big data.

Whether from the fields of life sciences, Engineering, Chemistry, Astronomy, or the Environment, a few key points were made clear:

  • There still remain problems relating to ownership and privacy of Big Data
  • Non big-data specialists face difficulties in analysing large volumes of data
  • Standards on storing and analysing vary by institution, regionally and nationally.
  • Big Data is going to play a huge role in all facets of science and engineering in immediate and further future.

The purpose of the conference was to find out ‘what are researchers’ biggest questions on Big Data’, and how we might come to resolve these in future.  The event invited established academics to talk specifically with early career researchers and workshop on a case-by-case basis. By showcasing their own research, academics were able to communicate effectively with ECRs to raise and explain potential research problems in managing big data.

Browse Event Gallery

From our guests

sinan

Guest Post: Sinan Shi talks #SESBigData

By | Big Data Guest Post, News, The Data Revolution | No Comments
Sinan Shi, Research Software Developer at University College London, was among the invited guests at SES's Big Data event. He recaps the day and reflects on the benefits of reaching out of researchers of different disciplines. Big data is probably the one of hottest terms being used in recent years. The scientific community is one of the first communities to become aware of the challenge and opportunity in big data. Data will play a more and more crucial role in the future of scientific research. In this workshop, instead of talking about the details of big data techniques or general big data...
Read More

Workshop Results

1
Expectations
2
Workflow

Managing academic expectations: researchers are having to get used to thinking of data access and processing as an added cost and very expensive resource. Costs of buying data and storage requires more transparency and openly available information. The amount of time required to complete datasets is well underestimated, leading to time-management issues for researchers and ECRs.

The ways we analyse and manage data workflows varies across teams and institutions. Every step of the way there is needs to be standards in:

  1. Storage
  2. Analysis
  3. Curation
  4. Maintenance
  5. Retrieval
3
Policy
4
Value
5
Access

There is still uncertainty in Research Council policies and how researchers can comply with them. E.g. what needs to be shared and how, what level of ‘granularity’ and metadata is deliverable, is raw data or analysed data required?

Who is to decide the value of data? We cannot store everything, so how do we know what might be useful for others even if it is not for us? This also raises issues of original intention, i.e. how are we able to find out what data is about if the researchers who created it are no longer available.

Data sharing across organisational boundaries leaves researchers very aware of the boundaries in data access, no only due to organisational dynamics but also the logistics of sharing large quantities of that data. There is a large discrepancy in access to data dependant on faculty and background, especially in regards to Applied Science vs. Social Science. Long term data accessibility is a persistent question, especially in terms of selecting hardware and data formats. Organisations need to rethink of access to data as a discipline in itself, requiring dedicated staff.

Catch up on talks

Discover More

Download Presentations/Video