Ground Truth
By George M. Garrity. 2009
120 citations recorded by the source corpus.
The concept of ground truth is a well-established principle in cartography, where data collected at a distance are confirmed by measurements made on location. Those local measurements are used to calibrate remote sensing devices, verify or correct experimental inferences, and update geographic databases. Ground truth observations also provide a means of training and supervising image classification software and resolving errors of omission or commission. Cartographic methods have improved significantly because of the development of precise positioning methods (GPS), the development of interoperable data standards for rapid exchange of precise and highly interlinked information , and the development of various devices and visualizations that serve-up information on-demand to different classes of end-users. There are strong parallels between mapping geographical and biological space. Much of what is underway in genomics, evolutionary biology and systems biology is analogous to the development of a coordinate system onto which living systems can be mapped, natural boundaries and interrelationships uncovered, and predictions of properties and behaviors based. However, it is likely that the dimensionality of any biological coordinate system will exceed the four dimensions of the geographical system. This will confound visualization and complicate " navigation " through biological space, whether it is for purely exploratory purposes or to get from one point to another. It is a given that the volume of biological data will continue to grow super-linearly for the foreseeable future, as new computational methods are applied to answer the " big questions " in biology. In the absence of major innovation, it is likely that the gap between the cost of data analysis and the cost of data generation will continue to widen. The outcome of such analyses are highly dependent on the quality of the input data, including the underlying information and knowledge used to inform the creation of datasets, the algorithms used in analyses, and the interpretation. Errors of commission and omission appear to be more common in biological data sets than physical data set and the former are more likely to be affected by semantic ambiguity and hidden biases. What is not yet established is which of the labor-intensive cu-ratorial and interpretive tasks can be automated and what metadata that is absent from the public databases may be located and recovered from other sources in a usable form. The impact of semantic ambiguity in biological data has been noted previously as it pertains to iden-tifiers [1] or biological names …
Subfields: Ecology, Evolution, Behavior and Systematics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Ecology, Evolution, Behavior and Systematics
Citations in the corpus
- Your fairness may vary: Pretrained language model fairness in toxic text classification (2021, methodology): “Motivated by this observation, we started looking into understanding the quality of datasets used in toxic text prediction (Arhin et al., 2021).”