Leveraging Data Science to Explore Co-Relationships Between Elements and Minerals
University of Idaho Seed Grant
The geoscience community has discovered more than 5000 mineral species so far. A topic of great interest is how the constituent elements parse and associate themselves into minerals during the Earth's evolution. This project aims to explore the co-relationships between the 72 mineral forming elements and the presently known mineral species. Data science methods, including data management, data analysis and data visualization will be applied to: (1) Retrieve and cleanse dataset from a mineral database mindat.org; (2) Conduct statistical analyses on numbers of mineral species containing 1, 2, and 3 elements as well as the change of numbers such as "Does the presence of element Z affect the correlation between elements X and Y in mineral species"; and (3) Develop new three-dimensional data visualization techniques to support the data analysis procedure and demonstrate the results. The proposed work requires cross-disciplinary knowledge between computer science, statistics and mineralogy, and will facilitate collaboration between the PI's group and other research units in- and outside the campus. Expect deliverables are a tool for visualizing and analyzing mineral-element correlations, publications introducing the research and findings, enhanced program of a data science course, and identified topics for proposals towards external funding.