OptionalbrowserMap of browser keyword (matched case-insensitively in the user-agent) →
frequency, e.g. { "chrome": 0.63, "safari": 0.19, "firefox": 0.04 }.
OptionalhardwareList of hardware profile tuples with associated frequency. The best-matching profile (exact concurrency + memory match) is used.
OptionallanguagesMap of primary language tag → frequency, e.g.
{ "en-US": 0.72, "en-GB": 0.08, "fr-FR": 0.05 }.
OptionalplatformsMap of platform string → frequency, e.g.
{ "Win32": 0.85, "MacIntel": 0.12 }.
OptionalresolutionsMap of "WIDTHxHEIGHT" resolution string → frequency, e.g.
{ "1920x1080": 0.55, "1366x768": 0.13 }.
Frequency table used to describe how common each fingerprint signal is in a specific user population.
All frequency values should be normalised to
[0, 1]where:1.0= present in every request in the population (maximum attractor risk)0.0= never seen in the population (no attractor risk)Omitting a dimension causes that dimension to fall back to the built-in heuristic for that signal.