Fighting the Illegal Trade in Endangered Plants Through a Cross-Disciplinary Approach
Updated: 9 hours ago
Dr Anita Lavorgna, Principle Investigator of the FloraGuard project, is Associate Professor in Criminology at the University of Southampton. Anita's research pivots around cybercrimes. Dr Stuart Middleton, Co-Investigator, is Lecturer in Computer Science at the University of Southampton.
Illegal commerce in plants and their derivatives--including may orchid species globally--is a market-based crime. It has been boosted by the commercialisation of the Internet, and threatens and destroys numerous species and important natural resources, and may cause phytosanitary and health problems. Over the years, this issue has received attention mostly in the area of conservation science and, more recently, also in the field of green criminology - a fast developing field of the social sciences focusing on the harms and crimes committed against the natural environment, including animals and plants living in it. Despite the consensus that a better understanding of the characteristics of online illegal markets, and of the actors operating in them, is a necessary starting point for any intervention to mitigate the problem, there are still several research gaps on the role of the internet as a facilitator in the illegal trade of plants, and on its policing.
With our UK Economic and Social Research Council-funded project FloraGuard (http://floraguard.org/), we combined cross-disciplinary ways of analysing online marketplaces for the illegal trade of endangered plants (bringing together criminology, plant ecology, computer science and law enforcement expertise) with analyses of existing policing practices to assist law enforcement in the detection and investigation of illegal trades of endangered plants. In this context, aiming to identify solutions for preventing the illegal trade in animals and plants, we created a software using artificial intelligence to extract data such as payment details, wildlife species and location, from hundreds of forums and marketplaces; it can also recognise recurring names and key words that might be used in illegal markets online. The information collected by the software is then converted into charts, diagrams and pictures which can be used in investigations or to trigger new ones.
More importantly, we laid the ground for improving cross-disciplinary ICT-enabled methodologies for investigating internet-enabled illegal activities, proposing an iterative approach which facilitates the coordination of different specialist expertise, as well as the coordinated use of qualitative and quantitative methods. We are aware, however, that as computational criminology and sociotechnical approaches are becoming increasingly common, issues related to "datification" are bound to increase as well. In the effort to use explainable and provenance-preserving AI techniques, we propose that in similar studies an “AI review” step should be included to review the wider sociotechnical system - that is, a systematic way to check for potential bias in sampling and training data that AI systems are using. In this way, it would be possible to make the final decision makers aware of any limitations in coverage behind online crawled data, in order to build trustworthiness into the AI approaches used.
Read the full paper here.