Outlier detection based on distribution fitness

A computerized method comprising:

• receiving, via a first computerized user interface implemented as a screen of a graphical user interface, a submission of a confidential data value of a first confidential data type from a first user, entered into a field of the screen of the graphical user interface;
• identifying, using one or more processors, one or more attributes of the first user;
• retrieving, using one or more processors, a plurality of previously submitted confidential data values of a first confidential data type for a cohort matching the one or more attributes of the first user, the previously submitted confidential data values having been encrypted on an external data source, the cohort being a grouping of data pertaining to a combination of user attributes for users who submitted the confidential data values;
• deriving one or more intermediate cohorts by generalizing each of the one or more attributes of the cohort up at least one level in a different taxonomy corresponding to each of the one or more attributes, each taxonomy being a hierarchical classification of attribute values for a particular attribute, with each level of each taxonomy indicating attribute values that are generalizations of child attribute values at a lower level;
• selecting one or more of the intermediate cohorts;
• fitting a parameterized distribution to the previously submitted confidential data values that are contained within the selected one or more of the intermediate cohorts, outputting one or more estimated parameters for each of the selected one or more of the intermediate cohorts;
• setting a lower limit for the first confidential data type based on the one or more estimated parameters;
• determining whether the confidential data value submitted by the first user is an outlier by determining if the confidential data value submitted by the user is lower than the lower limit;
• and
• in response to a determination that the confidential data value submitted by the first user is not an outlier, permitting the confidential value submitted by the first user from to be used for insights provided to other users.

A non-transitory machine-readable storage medium comprising instructions, which when implemented by one or more machines, cause the one or more machines to perform operations comprising:

• receiving, via a first computerized user interface implemented as a screen of a graphical user interface, a submission of a confidential data value of a first confidential data type from a first user, entered into a field of the screen of the graphical user interface;
• identifying, using one or more hardware processors, one or more attributes of the first user;
• retrieving, using one or more hardware processors, a plurality of previously submitted confidential data values of a first confidential data type for a cohort matching the one or more attributes of the first user, the previously submitted confidential data values having been encrypted on an external data source, the cohort being a grouping of data pertaining to a combination of user attributes for users who submitted the confidential data values;
• deriving one or more intermediate cohorts by generalizing each of the one or more attributes of the cohort up at least one level in a different taxonomy corresponding to each of the one or more attributes, each taxonomy being a hierarchical classification of attribute values for a particular attribute, with each level of each taxonomy indicating attribute values that are generalizations of child attribute values at a lower level;
• selecting one or more of the intermediate cohorts;
• fitting a parameterized distribution to the previously submitted confidential data values that are contained within the selected one or more of the intermediate cohorts, outputting one or more estimated parameters for each of the selected one or more of the intermediate cohorts;
• setting a lower limit for the first confidential data type based on the one or more estimated parameters;
• determining whether the confidential data value submitted by the first user is an outlier by determining if the confidential data value submitted by the user is lower than the lower limit;
• and
• in response to a determination that the confidential data value submitted by the first user is not an outlier, permitting the confidential value submitted by the first user from to be used for insights provided to other users.