BD.20.015 – AI for biodiversity – using AI to empower citizens scientist to measure and understand biodiversity

Route: Creating Value through responsible access to and use of big data

Cluster question: 112 Can we use Big Data and Big Data collection to define values, generate insights, and get answers?

The world faces a biodiversity crisis and in order to address this there is an increasing need for high quality data on the distribution and trend of plants and animals. Increasingly this data is collected by citizen scientists who submit their records to popular websites like Waarneming.nl. It is difficult to judge which records are of high enough quality to be applied in scientific studies. Recently AI-based image recognition has proved to be able to support the identification skills of volunteers. This led to a surge in the number of records, an increase in active volunteers and an increase of the quality of data. We will take this a step further by developing an educational AI that interacts with the user like a teacher resulting in a higher data quality and a user who is able to understand the AI’s predictions. Such an AI supports and empowers people instead of reducing them to docile ignorant users.
To support this interactive AI we will improve the taxonomic process of species identification by fusing current machine learning algorithms and knowledge systems. Bridging the gap between perceptual knowledge in current learning algorithms and conceptual knowledge systems is a fundamental and crucial challenge in AI.
This project will be beneficial to the field of AI and to the field of biodiversity studies. It results in tools that can be used for a wide range of (citizens) science projects, increases the amount of trustworthy data and makes AI an integral part of biodiversity studies. For the field of AI the projects allows to study how users interact with AI and to test methods aimed at improving the AI-human interaction. The methodology can be applied to other application domains which involve both a perceptual and a conceptual part, such as healthcare or inspections in manufacturing.

Keywords

Artificial Intelligence, Big Data, biodiversity, citizen science, human-machine interaction

Other organisations

HAS Hogeschool, Hogeschool NHL Stenden, JADS, Universiteit Twente (UT), Vrije Universiteit Amsterdam (VU)

Submitter

Organisation Naturalis Biodiversity Center
Name Dr. V.J. (Vincent) Kalkman
E-mail vincent.kalkman@naturalis.nl
Website https://www.naturalis.nl/vincent-kalkman