search

Aircloak / aircloak

This repository contains the Aircloak Air frontend as well as the code for our Cloak query and anonymization platform
2 stars 0 forks source link

Support min/max on strings in restricted non-anonymizing queries #4051

Closed sebastian closed 4 years ago

sebastian commented 4 years ago

We currently do not support min and max for strings. It's not at all obvious how one would go about deriving a noisy string value. However there are many places in a query where min or max of string values can be sensibly generated, namely in a restricted per-user query as well as in an unrestricted query.

At Zurich they have this specific use case/need.

An example would be:

SELECT
  ...
FROM (
  -- The NPS table contains multiple entries per
  -- user, and they want a single value instead.
  SELECT 
    uid,
    column,
    max(nps_value)
  FROM nps
  GROUP BY uid, column
) deduplicatedNpsScores
...

Atlassian ticket: https://aircloak.atlassian.net/browse/SUP-5

cristianberneanu commented 4 years ago

We currently do not support min and max for strings. It's not at all obvious how one would go about deriving a noisy string value. However there are many places in a query where min or max of string values can be sensibly generated, namely in a restricted per-user query as well as in an unrestricted query.

This is not true. I just checked and we do support min/max over text in restricted/standard queries.