[Turing-Southampton] FW: Dstl-funded expression of interest (Horizon Scanning)

[Susan Davies]

***apologies if you receive this more than once***

Please see below a request for expressions of interest for a 6-month project funded by Dstl. They are also open to proposals that require more time.


Overview

The Turing's Defence & Security(D&S) programme would like to request expressions of interest for a 6-month project on conducting research into and trials of techniques that could be used to semi-automate horizon scanning for D&S, funded by Defence Science and Technology Laboratory (Dstl). We are seeking a Principal Investigator (PI) to undertake this project, starting in January 2022 and concluding by June 2022. The funder will cover direct and indirect costs, plus overheads.


To express interest, please email Alex Harris at dsprogramme at turing.ac.uk <mailto:dsprogramme at turing.ac.uk> with the names of interested individuals, a short profile on suitability for the project, and CV.

Responses are requested with a deadline of 25th October 2021.


Background Detail


Within the Defence Science and Technology Futures (DSTF) programme, the Science and Technology (S&T) Intelligence Project seeks to utilise advances in data science and machine learning to automatically ingest, process, and collate the information feeds required to understand the shifting S&T landscape.


Objectives of the S&T Intelligence Project:

  *   Develop a best-in-class platform to allow users to explore, summarise, and investigate areas of S&T worldwide (covering internal and external data sets);
  *   Conduct advanced analytics on S&T data sources incorporated into the platform to enable semi- automated or automated horizon scanning.


Research Aims


The aim of this task is to conduct research into and trials of techniques and content for the early detection of emerging technologies or threats for D&S. A key challenge is identifying how to filter and prioritise source material that is relevant and significant to the advancement of S&T for D&S.


Task objectives:

     *   to have a set of plausibly implementable methods that flag technology emergence relevant and significant to D&S with as much coverage of the S&T horizon as possible;
     *   with as early an indication of emergence as possible;
     *   whilst providing a sustainable flow of information.

Methods:


An ensemble of methods is likely to be required to provide the best possible chance of discovering useful content from any given data. Methods that give an indication of why content has been selected would be preferred, as this will help in reviewing effectiveness and support analysts in assessing relevance.


Possible indications of relevant content for S&T advancement in D&S might include:

     *   Scientific publications predicted to be highly impactful in a relevant field;
     *   Content that suggests a potentially disruptive change in capability or contains significant claims

(e.g., "world's best", "significant improvement", "first ever");

     *   Content that contains emergent terms or previously unseen but growing combinations of terms/topics/disciplines;
     *   Changes in the dynamics of scientific collaboration (e.g., emergent collaborations between authors/institutions/disciplines or key players connecting networks of researchers);
     *   Research transitioning from basic to applied, patent generation, or venture capital investment.

Experimentation:


Testing is expected to be carried out on a subset of the vast amount of relevant data freely available on the internet, but researchers must be careful to make this data as representative as possible and not overlyrely on techniques which will not generalise across domains.


The test data is expected to be English language only. Full text or metadata can be used, but the research needs to be mindful of the ability to scale methods to very large collections of documents.


Collecting and curating data is not expected to be a focus of this activity - the priority is methods used and having a minimum viable test data set.


Outputs

Essential outputs:

  *   Review and expand the provided literature search and prioritise methods to try;
  *   Create new hypotheses and methods to fill gaps left by the available literature in how to automatically identify relevant horizon scanning content;
  *   Develop well-structured code to implement and generalise the methods identified or created and test them on an agreed data set.

Desirable outputs:

  *   Identify or develop test data/methods to allow comparison of the effectiveness of different approaches in identifying relevant source content;
  *   Reusable Python3.9 code and/or Terraform or Cloud formation (infrastructure as code) to deploy the methods trialled on Amazon Web Services (AWS).

__________________________________
Susan Davies
Coordination Manager, Web Science Institute<https://www.southampton.ac.uk/wsi/index.page?>
University Liaison Manager, The Alan Turing Institute<https://www.southampton.ac.uk/wsi/alan-turing-institute/alan-turing-institute.page>
Web Science Institute
University of Southampton
Southampton SO17 1BJ
M 07768 266464
https://www.southampton.ac.uk/wsi<https://www.southampton.ac.uk/wsi/index.page>

My working days are Monday to Thursday.





-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ecs.soton.ac.uk/pipermail/turing-southampton/attachments/20211006/28ace92d/attachment-0001.html