Why exclude some elements from your verbatim?

There are many reasons why you may want to exclude certain words or statements from your verbatim. For example, most of the time you will exclude personal data (bank details, phone numbers...), religious or insulting remarks etc. You may also need to exclude certain words or expressions that are specific to your business or industry.

How to manage these exclusions in MyFeelBack?

When a respondent writes the answer to an open question, the content is scanned by MyFeelBack directly into the browser. If the verbatim contains words, sequences or expressions that must be excluded, these are automatically replaced by generic terms (or patterns). For example, if the verbatim contains a URL, it will be replaced by [**WEBSITE**]. Thus, the data originally written is never collected or stored.

The "Verbatim desensitization" option is available for all "text field" type questions in the "Advanced settings" tab. By default, it is deactivated:

When the option is active, verbatim is collected without accents and in lowercase. Excluded sequences are replaced by the pattern name in square brackets :

This option is compatible with the "Verbatim Analysis" option in MyFeelBack. In this case, the verbatim is analyzed without the excluded sequence. 

The following items are systematically excluded by MyFeelBack :

  • Email
  • Website
  • Dates
  • License plates
  • Passport number
  • Credit card number
  • Bank account number
  • CNI number
  • GNI
  • IBAN
  • BIC
  • Social Security Number
  • Sensitive keywords
  • Customer ID (a list of 8 consecutive digits)

How to add exclusions specific to my business?

MyFeelBack provides an editable list of excluded words to which you can add or substitute items.
To view it, go to the "Settings" tab and then under "Campaigns" click on "Verbatim Desensitization".

To add words to exclude, enter them directly into the list. Don't forget to return to the line for each new word or expression:

Notes : 

  • As all words are made lowercase and accent free, accents and case do not matter. For example: "diabete" and "Diabete" will be identified as only one word.
  • Plural and gender are not supported, so these versions should be added if necessary. For example: "house" and "houses" are two different words.
  • The words listed by default cover general themes (health, religion, sex, politics, origin...). 
  • To delete your changes and return to the initial list, you can click on "Reset".
Did this answer your question?