We invite Turing researchers (Turing Fellows, Turing Research Fellows, Doctoral Student Supervisors and others) and other academics from our partner universities to submit proposals for short projects to be funded through the Turing’s Defence and Security programme; a strategic partnership of The Alan Turing Institute with GCHQ and MOD (Dstl and JFC) on data science and AI research.
The short projects should be ~6 months long, start around October 2018, and finish by 28 March 2019.
The challenges from the programme under which this call is placed are social data science and privacy and trust - specifically digital identity assurance and data synthesis.
Further information on the programme challenges and goals, existing programme projects, can be found online here.
Full details are available in the call documentation; a summary is presented below.
Deadline for proposals: 28 August 2018 16:00
A total of £150k will be available, to be awarded for two or three projects. These should be short pilot/pump-priming projects, which can begin in October 2018 and will finish and have reports submitted by 28 March 2019.
These awards may act as a “stage gate”, with the potential for larger amounts of follow-on funding available in 2019-2020 for projects which have deliverables with potential for high impact and benefits.
Projects will also have access to Programme contacts, industry expertise and meeting space at the Turing HQ in London.
The funding can be used for direct and indirect costs (up to 80% fec). More information on eligible costs is available in the call documentation.
This call is seeking proposals for projects that apply techniques within the field of data science applied to digital identity assurance and data synthesis.
It is not expected that proposals address any more than one of the areas below (i.e. one bullet point).
Digital identity assurance
The UK Government has a mandate to protect its citizens at home and abroad; this includes aspects of their privacy and security as consumers of digital services.
The topics of interest for this research call are:
- How can data science help to provide greater assurance of identity during online interactions without compromising the anonymity of participants?
- What are the ways of spotting anomalies within homogeneous populations? (i.e. spotting an adult on a teenage group chat) – Can you spot “illegitimate” efforts to assimilate within a group?
- What techniques (and indicators in data) can be used to provide an assurance of identity? How normal are “normal” people within a given subgroup? How can we quantitatively contrast natural entropy with anomalies?
- There are scenarios in which people may wish to legitimately obfuscate their identity but remain digitally active – witness protection, child-protection and relocation, armed forces serving overseas. In such cases, there may be a need to protect digital users from a discovery that could directly impact their safety. Are there data-driven approaches to:
- Determining whether an online profile is receiving “unusual” attention? For example, how many of someone’s “Facebook friends” are genuine friends or connections?
- Determining whether a profile has been cloned or taken over by another user?
The sensitive nature of personal data makes sharing notoriously difficult for governments (and indeed many other types of industry) to do effectively. Often, proxy datasets are sought (which is not ideal) or datasets are anonymised to the point where any interesting features have been obfuscated.
- What methods exist for generating representative data that carries the same data characteristics as the training data, yet carries none of the same identifying features? Is it possible to synthesise data characteristics to yield translational findings to the original data?
- How does this extend to synthesising human behaviour (i.e. synthesising events around a particular narrative)? How effectively can we generate activity logs, location or social trajectories?
How to apply
Applications must be submitted via the online portal on Flexi-Grant. If you have not already done so, all applicants must first register on the system and provide basic details to create a profile. If you have any question regarding the application form or using the online system, contact the Programme Manager or Research Administrator via [email protected]
Deadline: 28 August 2018 16.00
Principal Investigators (PIs) should be Turing researchers or researchers affiliated with one of the partner universities. Collaborators on projects are welcome from anywhere outside the Turing or the partner universities.
- Publication of call for proposals: 6 August 2018
- Q&A sessions: 15 August 2018, 14:30-17:30
- Deadline to submit: 16:00 on Tuesday 28 August 2018
- Eligibility checking, resubmission requests: 27 – 30 August
- Review and selection of proposals: Early-mid September 2018
- Award offers and acceptance of terms: Mid-late September 2018
- Projects start: From 1 October 2018
- Projects to complete and report by: 28 March 2019 at the latest
We encourage pre-submission discussion on the scientific direction of your proposal with the Programme Director, Mark Briers, [email protected] (and Programme Visiting Researcher Ben Tagger, [email protected], after mid-August)
Any queries regarding the process, post-award requirements, costing or general eligibility should be discussed with the Programme Manager, Catherine Lawrence, and/or Research Administrator, Charlie Thomas (also on [email protected])
For more detailed queries on the research elements, potential applicants can book a meeting or phone call with representatives of the funders and the Turing on 15 August between 14:30 and 17:30.
Representatives will be available for 25-minute appointments. To book your slot, please complete the form here: In the ‘Your name’ box, please enter your name along with F2F for in-person meeting, TC for a telephone call or S/Z for via Skype or Zoom.
If you are not able to book a slot on 15 August, please email [email protected] and we will try to book you a time on another day.
Charlie Thomas, Research Administrator
Catherine Lawrence, Programme Manager / 0203 862 3381
Mark Briers, Programme Director