Data mining
Hunting for patterns in large datasets — a term statisticians originally used as an insult.
The phrase came out of the database community around 1990. Among statisticians, "data dredging" had long meant searching until something looked significant, which in a large enough dataset always eventually happens. The field responded with methodology: the KDD process and later CRISP-DM (1999), which forces the work into phases from business understanding to deployment, and where most of the time goes on cleaning data.
The classic techniques are association rules (Apriori, 1994 — market basket analysis, and the endlessly retold and never substantiated story about nappies and beer), classification, clustering and anomaly detection. Since 2012 the label has largely been replaced by data science or machine learning, but the toolkit is broadly the same, and so is the underlying failure mode of finding patterns that do not generalise.