PhD: Identification of internet-based illegal wildlife trade through machine learning

Vice Chancellor’s Research Scholarship
International Wildlife Trade
Durrell Institute of Conservation and Ecology School of Anthropology and Conservation

Project title: Identification of internet-based illegal wildlife trade through machine learning

Project description
The trade in wildlife, including animals, plants, their parts and derivatives involves markets as diverse as fisheries, timber, medicines, foods and exotic pets. However, illegal wildlife trade can lead to the extinction
of traded species and frequently impacts on non-targeted species through ‘by-catch’ and can also help spread infectious diseases. It is particularly pernicious as it widens corruption, fuels conflict and hinders development. Environmental crime, including the illegal wildlife trade, is now estimated to be worth $91-258 billion p.a., making it the fourth most valuable form of crime after narcotics, counterfeiting and human trafficking.

Global online trade, including in wildlife, is growing. It allows small business to prosper and reach a global audience; however, it also masks increasing numbers of illegal transactions. It is clear that, with the increased publicity on the illegal wildlife trade and a push to enforce and expand current local and national legislations, illegal wildlife trade is increasingly moving online. However, the identification of illegally traded items can be challenging and is largely undertaken by law enforcement manually, scrolling through pages
and inspecting each individual item.

This project aims to develop novel insights into the detection of illegal wildlife trade being sold over the internet. Specifically, the student will employ innovative algorithms based on machine learning (and potentially Deep Learning) to identify species for sale from images and the metadata associated with online adverts. This will allow for the automated identification and reporting of illegal wildlife trade on sale via online marketplaces, social networks and the darkweb. The results of this project will be of particular
interest to law enforcement and governments attempting to curb these and other illegal activities.

Eligibility Criteria

The successful candidate will

  • have a good Honours degree (First or 2i) or a Master’s degree at merit or distinction in the areas of Bioinformatics, Computing, Engineering, Mathematics, Conservation Science or a related subject,
  • be able to demonstrate experience in computer programming, and
  • have experience in machine learning, data mining or deep learning.

It is desirable that the candidate has experience in

  • Python, Java, C, C++ programming languages
  • web development
  • image processing
  • interacting with social networks via APIs
  • UK, EU and overseas fee paying students are invited to apply. Please note that overseas students must have the appropriate documentation to evidence eligibility to work in the UK

Further information

Start date: 15 September 2018
Programme: PhD Biodiversity Management
Mode of study: Full-time
Studentship length: 3 years
Application deadline: 23.59pm Wednesday 28 February
£14,553 (2017/18 rate) plus Home/EU fees
Scholarships will be offered at the standard UK Research Councils’ rate and administered under the Graduate Teaching Assistant Scheme, further details of which can be found at:

How to Apply
Applicants should submit an application for the PhD Biodiversity Management via the Kent website:

Please copy the research project title and description into the appropriate boxes on the application form, a separate proposal is not required

Applications should include certificates and transcripts for all completed and pending degree level qualifications, a CV detailing relevant experience, and the official contact details for two referees

Contact Details
For queries regarding the project or eligibility please contact Dr David Roberts
For queries relating to the online application form please contact Rebekah McComas