Data quality principles applied to agricultural research
Year: 2017
Type: other
Abstract: Data-driven decision making has become a reality in the context of farming practices worldwide. On the other hand, our hability to effectively promote data reuse is constrained by our capacity to interpret it, especially across disciplines. Agricultural research data arise from different disciplinary fields and are generated from heterogeneous sources and techniques, such as ground measurements and remotely sensed data, bringing together challenges associated with data quality control and assurance. However, although it is agreed that high quality data are a precondition for better data driven decision making, quality standards and quality assessment methods are lacking. We describe our effort on building a strategy to apply data quality assessment principles within Embrapa, the Brazilian Agricultural Research Corporation. Embrapa is a networked public company composed by 46 research centers distributed throughout the country, where it conducts actions to promote solutions on research, development and innovation for sustainable agriculture. We propose to discuss the following principles in order to assure agriculture data quality within its life-cycle management: discoverability, accessibility, usability, reliability, relevance, preservation and curation. In terms of scientific quality, such principles are linked to aspects of accuracy, precision, uncertainty, validity and fitness for use, the latter being related to the perception of relevance and capacity to interpret the data by user communities. We argue that well documented metadata stored in reliable repositories are one key factor to enable data quality assessment properly and we provide insights from our experience towards improving the quality of agricultural datasets.
Institution: Embrapa
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
SDG: Zero hunger
Open Access status: gold
OpenAlex ID: W2891685348