PhD Kulkarni Ritwik

Addressing the global biodiversity crisis in the Information Age: machine learning methods and online data

Tieteellinen tutkimus / siihen pohjautuva työ | Nelivuotinen

One of the greatest sustainability challenges facing our society is the loss of biodiversity at a rapid rate. Wildlife trade is one of the greatest threats to biodiversity and it also affects humans because of the links with zoonotic diseases, such as the SARS-CoV-2 virus . With the advent of digital marketplaces and social media, the trade has gone online, and in many cases, it violates laws for protected species. Reports suggest that the covid-19 pandemic has contributed to a significant rise in the online trade of threatened species. In this project I will develop automated methods using machine learning to identify and analyse the trade on digital medium. The work will focus on two main modalities of digital data, namely, images and text. Thus, building neural network models for machine vision and natural language processing. Machine vision models will address two aspects i) detecting images in the context of sale of wild species and ii) identifying specific species to assess the present and future risk for potential population decline. Text processing models will similarly tackle i) identifying and extracting species specific information from posts which advertise selling of exotic animals and plants and ii) identify patterns and extract information from a broader analysis, related to prices, quantities and trade routes for wildlife trade. The outcomes of the project will result in building novel methodologies and datasets in relation to automatic identification of wildlife trade related content in the digital space. Thus, assisting stakeholders across research communities, international organisations and law enforcement agencies to help stem the global biodiversity crisis.