The utility of a human/computer learning network for improving biodiversity conservation in eBird

Опубликовано 12 августа 2016, 2:57

We describe our work to improve the quality and utility of citizen science contributions to eBird, arguably the largest biodiversity data collection project in existence. Citizen science (the use of human sensors') is especially important in a number of observation-based fields, such as astronomy, ecology, and ornithology, where the scale and geographic distribution of phenomena to be observed far exceeds the capabilities of the established research community. Our work is based on the notion of a Human/Computer Learning Network, in which the benefits of active learning (in both the machine learning sense and human learning sense) are cyclically fed back among human and computational participants.'