MetaMatchMaker
For researchers who need to find and integrate data quickly and get to the science faster
Pretrained Learning Model
Creating a graph database of common data elements using transfer learning. Getting scientists to analysis more quickly by automating data searching and integration.
In the age of big data, the data generation vastly outpaces analysis. There are mountains of data waiting to be mined for new discoveries. But data integration and harmonization are time consuming and expensive. MetaMatchMaker helps scientists find data and automate common data element mapping so they can get to analysis quickly.
- Quickly search for new data by keyword
- Identify compatible data sets
- Model-based searches of natural language phrases
- Summary statistics of data (e.g., sample size, demographics)
NSF Convergence Accelerator
This work is funded by the NSF convergence accelerator program, a new approach to bringing experts from different backgrounds together to solve big problems in science.
MetaMatchMaker is part of the 2020 Cohort: Delivering societal impact through quantum and AI-driven data and model sharing research.
Partners
This project could not work without an incredible team of experts in bioinformatics, statistics, machine learning, and data harmonization.