This project uses machine learning to automatically detect important details from research papers, such as when a specific dataset or software tool is mentioned, what scientific instruments were used, or which grant supported the work. These connections are then used to improve search results, highlight related work, and help researchers more easily understand how different pieces of science are linked.

Originally developed for astrophysics, SciX is now expanding to support Earth science. This includes making it easier to find papers that use NASA Earth data, match studies to geographic locations or environmental phenomena, and connect publications to Earth science vocabularies and data repositories. The goal is to make it faster and easier for Earth scientists to do research, while also helping data centers, funders, and the public see how data and tools are being used.