Collaborative Research: Research Infrastructure: Biodiversity On Demand (BOND): A Web Platform for Discovering, Comparing, and Integrating Biodiversity Databases

A grand challenge in this era of rapid global change is efficiently assessing, predicting, and ameliorating changes in biodiversity and associated ecosystem services. Addressing this challenge requires unprecedented access to biodiversity data across spatial, temporal, and taxonomic scales. The past few decades have witnessed the rapid growth of biodiversity information. These data comprise, but are not limited to, information on species’ geographic distributions, functional traits, phylogenetic relationships, phenology, and biotic interactions. However, these data are scattered across numerous, often isolated databases and repositories, which have proliferated largely in parallel. This makes searching and synthesizing relevant information difficult, limiting our ability to effectively harness the entirety of available data and identify gaps in our knowledge. As a result, it has become difficult to assess what data we already have and where to find them. Closing these gaps is necessary for understanding and managing biodiversity loss, which affects human health, food security, and the economy. Therefore, the scientific community has been calling for greater alignment between efforts to achieve a more complete picture of Earth’s biodiversity. This project will construct Biodiversity on Demand (BOND), a pilot web platform for discovering, comparing, and integrating biodiversity databases. BOND is expected to greatly enhance the community’s capacity in database discovery, comparison, and integration, thus helping the science community to more effectively use the rapidly growing volume of biodiversity data to gain a more complete picture of Earth’s biodiversity. The project will promote the training of the next generation of scientists by providing support for undergraduate students through the Research Experience for Undergraduates Site at Emory University and graduate students at the institutions of the project team. Students will gain cross-disciplinary training in biology and computer science, especially gaining knowledge of big data and artificial intelligence, which are critical techniques in the era of biotechnological revolution. The project will actively involve participants from the global biodiversity informatics community and promote the long-term development of biodiversity informatics. The research team will work with stakeholders to co-design a metadata schema with maximized compatibility for indexing heterogeneous biodiversity databases. In addition to indexing, the metadata schema will also enable database comparison and integration. Building upon the metadata schema, the team will build a database of biodiversity databases to enhance the discovery and comparison of biodiversity data. The compiled metadata will be transformed into a distance matrix representing the similarities and differences among databases. The project will also develop deep learning algorithms to facilitate the integration of biodiversity databases, contributing new artificial intelligence approaches for synthesizing heterogeneous scientific data. The algorithms will be evaluated with a few representative case studies. Lastly, those functionalities will be integrated in a web interface. The long-term goal of BOND is to become the “first stop” for discovering, evaluating, and synthesizing biodiversity data. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.