NTEP‐DB 1.0: A relational database for the national turfgrass evaluation program

Abstract

AbstractGiven field experiment data collected by the National Turfgrass Evaluation Program (NTEP), we aim to design and create a relational database to store the data and support efficient queries. As one of the most widely‐known turfgrass research programs in the world, NTEP has generated large volumes of data on turfgrass cultivars and experimental germplasm since the early 1980s, providing invaluable information for a variety of user groups (e.g., homeowners, seed companies, golf course managers, retailers, turfgrass researchers) to select cultivars that best fit their needs (e.g., winter survival, pest tolerance, turf quality). The datasets have historically been stored in large sets of text files and spreadsheets. Currently, NTEP data are delivered to users through a website (www.ntep.org) as summary reports and it can be extremely tedious (e.g., hundreds of clicks, data merging, jargon) to perform a simple query (e.g., best cultivar selection with typical conditions). This significantly limits the use of NTEP data and hides its value from the public. To address these limitations, we carried out an interdisciplinary effort with horticulture and computer science researchers to design and create the first NTEP database – NTEP‐DB 1.0 – to reduce the manual efforts and expert knowledge currently required to extract meaningful information from the data. Experiments confirm that the query results are correct, and that the database can greatly reduce manual efforts. Anticipating next‐generation advances, we also recommend incorporating spatial data types and analytical techniques into future designs of the database.