Identifying information in resource locators

A system for identifying Uniform Resource Locator (URL) parameters that indicate personally identifiable information, the system comprising:

• computer memory for storing session records of a plurality of users, said session records specifying URL requests submitted by user devices to at least one server in at least one network domain;
• one or more processors configured to:
• identify a URL parameter that appears in URL requests of a plurality of the session records, said URL parameter represented in URLs as a key-value pair having a fixed key portion and a variable value portion;
• generate, for the URL parameter, a measure of uniqueness of the variable value portion across multiple sessions;
• and
• determine, based at least partly on the measure of uniqueness, whether to treat the URL parameter as being associated with personally identifiable information.

Non-transitory computer storage having stored thereon instructions that, when executed by a computer system having computer storage, cause the computer system to perform operations comprising:

• accessing session records of a plurality of users, said session records specifying Uniform Resource Locator (URL) requests submitted by user devices to at least one server in at least one network domain;
• accessing a data structure identifying URL parameters that are deemed to be associated with personally identifiable information, the URL parameters represented in URLs as key-value pairs having a fixed key portion and a variable value portion, the URL parameters identified based at least partly on a measure of uniqueness of the variable value portion across multiple sessions;
• and
• removing the URL parameters from at least some of the session records.