Semantic Data Integration For Forest Observatory
Project Overview
Poaching and animal trafficking are significant challenges around the world. Anti-poaching efforts are always underfunded and under-resourced, and law enforcement officers cannot keep up with the large number of poachers trying to kill and capture animals. Due to limited manpower, they cannot patrol and protect vast areas of land. This project semantically integrates data gathered by bio-science researchers and environmental scientists to predict where poaching activities will occur in the future. Data-driven prediction models identify areas and time frames that are highly likely to have poaching incidents, enabling law enforcement agencies to deploy their limited resources more effectively. The project focuses on the Lower Kinabatangan Wildlife Sanctuary in Sabah, Malaysia, and is a collaboration between the School of Computer Science and the School of Biosciences at Cardiff University, including its Danau Girang Field Centre.
The approach develops a Forest Observatory built on linked data and predictive analytics. A Linked Datastore integrates heterogeneous data from multiple sources, and data-driven predictive analytics support conservation decision-making by identifying where and when poaching activity is most likely to occur.
The system draws on both new IoT sensing infrastructure and a decade of existing datasets. While an Internet of Things infrastructure will be deployed for real-time monitoring, the project also utilises datasets already collected over the past decade, including animal collar data, camera traps, satellite imagery, LiDAR, and environmental data, each generated using different time frames, durations, and geographic areas. Semantically integrating these diverse sources enables richer predictive models than any single dataset could provide.
Team
Funding
