Advancing IceCube through Citizen Science, Machine Learning, and Physics Education Research

Name that Neutrino (NTN) is a thriving IceCube citizen-science project on Zooniverse with more than 5,100 registered volunteers. Since launching in 2023, this community has already generated over 460,000 neutrino classifications. IceCube is a massive neutrino telescope at the South Pole that records an enormous number of events to search for astronomical neutrino sources. NTN volunteers are asked to classify the shape of the IceCube event using pattern recognition. The goal is to increase IceCube’s sensitivity to neutrino astrophysics using the innovative approach of combining citizen science and machine learning, while also studying how the citizen science participants use and learn from the process of classifying events. A new partnership between the IceCube project and the Drexel Physics Education Research (PER) Network will capitalize on synergies - providing novel educational data while also improving neutrino characterization through improvements to Deep Neural Networks (DNN). NTN has identified a phase space of events where citizen science produces better classification compared to Deep Neural Networks (DNNs). Accurate topological classification is crucial for down-stream analysis given that different reconstruction algorithms are run based on the classification. Running an incorrect algorithm can cause events to be lost, or worse yet, be falsely classified as “good events” in further signal processing. This can then be retained tin final data sets contributing to background noise. Through this project, NTN will be used instead of a DNN to classify these events, and the new classifications will be used to better train new and improved DNNs. 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.