ERC Team Builds Novel Database Improving Data Analysis of Silicon Nanostructures
Outcome/Accomplishment
The Nanomanufacturing Systems Center (NASCENT), an NSF-funded Engineering Research Center (ERC) headquartered at the University of Texas at Austin, has built a database system based on a hierarchical tree structure. The novel organization structure increases both the search speed and accuracy of very large datasets obtained from silicon nanostructures.
Impact/Benefits
The search speed using the tree-based database organization was 672 times faster than the search speed using a traditional list-based database. Search precision of the data was 99.67 percent accurate, and recall measures – the ability of the system to present all relevant items – were 98.6 percent.
Explanation/Background
Nanostructures are made from a variety of materials and have dimensions of just a few nanometers. Silicon nanostructures show great promise for the improvement and development of biomedical and other sensing devices. With a large capacity to reflect light, nanostructures are measured by shining a light on the structure and measuring the light reflected. The resulting reflectance measurements generate enormous datasets that must be managed and mined to identify patterns.
In order to organize and utilize this data, Center researchers developed a data organization model –called a growing self-organizing map (GSOM) – based on a hierarchical tree-based structure. This model significantly increased both search speed and accuracy using just 10 percent of the data obtained from the silicon nanostructure. This model shows promise for continuous improvement with a corresponding increase in the dataset.
Location
Austin, Texaswebsite
Start Year
Advanced Manufacturing
Lead Institution
Core Partners
Fact Sheet
Outcome/Accomplishment
The Nanomanufacturing Systems Center (NASCENT), an NSF-funded Engineering Research Center (ERC) headquartered at the University of Texas at Austin, has built a database system based on a hierarchical tree structure. The novel organization structure increases both the search speed and accuracy of very large datasets obtained from silicon nanostructures.
Location
Austin, Texaswebsite
Start Year
Advanced Manufacturing
Lead Institution
Core Partners
Fact Sheet
Impact/benefits
The search speed using the tree-based database organization was 672 times faster than the search speed using a traditional list-based database. Search precision of the data was 99.67 percent accurate, and recall measures – the ability of the system to present all relevant items – were 98.6 percent.
Explanation/Background
Nanostructures are made from a variety of materials and have dimensions of just a few nanometers. Silicon nanostructures show great promise for the improvement and development of biomedical and other sensing devices. With a large capacity to reflect light, nanostructures are measured by shining a light on the structure and measuring the light reflected. The resulting reflectance measurements generate enormous datasets that must be managed and mined to identify patterns.
In order to organize and utilize this data, Center researchers developed a data organization model –called a growing self-organizing map (GSOM) – based on a hierarchical tree-based structure. This model significantly increased both search speed and accuracy using just 10 percent of the data obtained from the silicon nanostructure. This model shows promise for continuous improvement with a corresponding increase in the dataset.