시민과학에서의 AI 활용 가능성과 한계

초록

This paper analyzes how AI has impacted citizen science through case studies of its use in citizen science. The application of machine learning, deep learning algorithms, such as computer vision and convolutional neural networks, to citizen science has increased the speed of data processing and improved data quality produced by citizen science projects. The automation of data processing with machine learning algorithms has made it possible to track global changes in biodiversity and contribute to evidence-based, proactive conservation policy. The ability to provide personalized feedback and training to citizen scientists has led to increased participation in citizen science projects. On the other hand, AI technologies can also automate the scientific activities that citizen scientists need to perform, leading to situations where citizen science can be done without the participation of citizen scientists. The black box and bias issues of current AI technologies can also limit the possibilities of citizen science that can be contributed to constitution of another science. In order to sustain the positive effects that AI technologies can bring, it is necessary to develop AI use strategies that prevent the negative consequences and enable collaborative activities between citizen scientists and AI. We need to develop AI technologies as collaborative partners that ensure the participation of diverse citizen scientists.

키워드