DISK
DISK is a novel framework to test and revise hypotheses based on automatic analysis of scientific data repositories that grow over time. Given an input hypothesis, DISK is able to search for appropriate data to test it and revise it accordingly, and does this continuously as new data be-comes available. DISK is also capable of triggering new kinds of analyses when new kinds of data become available. The provenance of the revised hypotheses is recorded, with all the details of the analyses. Ongoing research includes extending DISK to generate interactive explanations for scientists based on provenance records, developing a general approach to the design of meta-workflows, handling more complex hypotheses, and exploring the use of this approach in other areas of science.
Links:
Repositories
Varun Ratnakar
Research Engineer
Varun Ratnakar is director of Karya Limited, which provides services in AI research and development to USC’s Information Sciences Instute.
Yolanda Gil
Senior Director for Major Strategic AI and Data Science Initiatives
I have broad research interests in AI and data science. As Principal Scientist I lead the Interactive Knowledge Capture research group, which is part of AI@ISI.
Daniel Garijo
Collaborator
Deborah Khider
Research Lead
My research interests lie at the intersection of geoscience and artificial intelligence. I am particularly interested in using AI to advance the field of paleoclimatology, including tools to annotate and retrieve data, workflows for analysis and machine learning for climate-relevant predictions.
Maximiliano Osorio
Research Engineer
I’m a software engineer who is passionate about making AI researchers possible, creating technologies to integrate information and models across disciplines.