Visit UNICEF Global
Data quality is a critical concern for AI projects. Poor data quality can bias prediction results and mislead users. There are six key areas of data quality:
Accuracy: how well does the data reflect reality?
Completeness: is the data comprehensive or not missing value entries in unexpected instances?
Consistency: does information stored in one place match the information stored in another place?
Timeliness: is data available when needed?
Validity: is data in a specific format? Is it an unusable format? Does it follow business rules?
Uniqueness: Is the data instance the only instance in which the data appears in the sample?
Sarfin, R. L., & Editor, P. (2021, May 12). Data Quality Dimensions: How do you measure up? (+ free scorecard). Precisely. Retrieved from https://www.precisely.com/blog/data-quality/data-quality-dimensions-measure
Updated on 19 Apr 2022