Data Anonymization
Lumoa has the ability to anonymize certain data as it passes through our systems. This guide will walk you through some of the details of that process.
Before you begin
🔑 Permissions: Please note that anonymization is a paid feature and will not be enabled by default in most cases. You need to contact your CS manager or Lumoa Customer Service to turn on anonymization.
When does anonymization happen?
When a feedback gets passed into Lumoa, our Analytics begins working on it. This has a number of steps associated with it, depending on your collection settings and the type of data you are uploading. We are going to cover the simplified required steps below:
Data gets submitted
Data is checked for correct tagging and naming (every feedback needs a "date", for example)
Data is anonymized
Data is translated into chosen language (default English)
Data has sentiment applied
Data is assigned into Topics
Data visible in UI
As you can see from the steps above, anonymization is done BEFORE the translation process, and NOT on your translated version of the feedback. Note that if your analytics are in English, and you submit feedback in English, the system will be able to determine that and it will not translate the feedback.
Here are those steps in a bit more detail:
Note: If you are are using the "translatedComment" variable in your data, then anonymization will not occur as the feedback is not being translated. If you are not using this variable, you can ignore this. Read more about this in our guide on uploading data.
What parts of my data will get anonymized?
The feedback anonymization looks for:
Number/letter strings of three or more
Such as phone numbers, SSN, ticket IDs, etc.
This includes numbers spelled as words, such as "one two three four"
Also includes separation like "one, two, three, four"
Only available for English, Nordics, and German languages
Email addresses
Such as proper host names emails such as [email protected], emails in the mailto:<email> pattern, but also precise email-like patterns.
Common first and last names for Finland, Sweden, Norway, Denmark, Iceland, England, Germany, Spain and Estonia.
Also, fully capitalized names like BOB and TERO.
Common street names/addresses for Finnish, English, Swedish, and Nordics.
With anonymization is turned on, any part of your customers' open text feedback that matches the above criteria will be replaced with ****** or <PERSON> instead of the actual text. After a feedback has been anonymized, it cannot be undone. Anonymization can be turned on using a per collection basis. Learn what is a Collection.
An example of an anonymized feedback can be found below:
Defining your own Anonymization Rules
In addition to the default rules above, Lumoa has the ability to create specific anonymization rules and triggers that can be custom made for your data.
With these rules, you could create custom anonymization rules for things like:
Anonymize e.g. product names ("Donald Duck") or name-looking Jargon ("BOB-report")
Sensitive customer information (license numbers, passwords, street names)
...anything that you wouldn't want all users in Lumoa to see (e.g. rare names passing anonymization filter)
These rules also allow you to create specific words and phrases that should or should not be anonymized. For example, this could be used to anonymize all names except those that are on your support team.
It's possible to make custom rules based on phrase matches, such as ("Walt Disney"), lemma matches (match "Bob" and "Bob's"), exact text matches ("BARBIE"), regular expression matches (car licenses like "ABC-123") or based on word context ("Password is X", "Regards, X")
Note that these rules will be created and maintained on Lumoa's side, with your input, should you choose to use them.
On the Collections page, under Anonymization, you can see which Collections are anonymized.
On = anonymized
Off = not anonymized
Get in touch
📧 Do you have any questions or comments about using Lumoa? Please don't hesitate to email us at [email protected].
SEO Keywords
Anonymisoi Kaikki Data