Other
Scientific paper
Jan 2011
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2011aas...21733402d&link_type=abstract
American Astronomical Society, AAS Meeting #217, #334.02; Bulletin of the American Astronomical Society, Vol. 43, 2011
Other
Scientific paper
Synoptic sky surveys are now discovering tens to hundreds of transient events every clear night, and that data rate is expected to increase dramatically as we move towards the LSST. A key problem is classification of transients, which determines their scientific interest and possible follow-up. Some of the relevant information is contextual, and easily recognizable by humans looking at images, but it is very hard to encode in the data pipelines. Crowdsourcing (aka Citizen Science) provides one possible way to gather such information. SkyDiscovery.org is a website that allows experts and citizen science enthusiasts to work together and share information in a collaborative scientific discovery environment. Currently there are two projects running on the website. In the Event Classification project users help finding candidate transients through a series of questions related to the images shown. Event classification depends very much form the contextual information and humans are remarkably effective at recognizing noise in incomplete heterogeneous data and figuring out which contextual information is important. In the SNHunt project users are requested to look for new objects appearing on images of galaxies taken by the Catalina Real-time Transient Survey, in order to find all the supernovae occurring in nearby bright galaxies. Images are served alongside with other tools that can help the discovery. A multi level approach allows the complexity of the interface to be tailored to the expertise level of the user. An entry level user can just review images and validate events as being real, while a more advanced user would be able to interact with the data associated to an event. The data gathered will not be only analyzed and used directly for some specific science project, but also to train well-defined algorithms to be used in automating such data analysis in the future.
Djorgovski Stanislav G.
Donalek Ciro
Drake Andrew J.
Fang Kejie
Graham Mary Jane
No associations
LandOfFree
SkyDiscovery: Humans and Machines Working Together does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with SkyDiscovery: Humans and Machines Working Together, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and SkyDiscovery: Humans and Machines Working Together will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1401663