Zipf's Law
Definition
Zipf's law describes the approximate relationship between word frequency and frequency rank.
An empirical rule according to which word frequency is approximately inversely proportional to its rank.
Simply explained
Few words occur very often, very many occur rarely. Frequency decreases roughly inversely with rank.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A search team uses the distribution for planning index and cache, not as an exact natural law.
Limits and pitfalls
- Exponent and valid range vary; log-log lines alone do not prove Zipf distribution or cause.
Origin & history
George Kingsley Zipf popularized the relationship in 1935/1949 in language and social data.
Use in mid-sized companies
- Possible application: A search team uses the distribution for planning index and cache, not as an exact natural law.
Frequently asked questions
What does Zipf's law mean?
An empirical rule according to which word frequency is approximately inversely proportional to its rank. What should you pay attention to? Exponent and valid range vary; log-log lines alone do not prove Zipf distribution or cause.
Sources
- Human Behavior and the Principle of Least Effort · Zipf, Human Behavior and the Principle of Least Effort
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: