The UI of the anonymizer looks like the following. The pink colored highlights are the mask. One can remove the mask just by right clicking on them. If you unmask the highlight, a grey highlight will be visible (like the "ui" word in the figure). If you click on the grey highlight, the earlier pink highlight will reappear. The pink highlights correspond to the permanent tags.
To include a new token sequence in your mapping from token sequence to entity tags, just select the sequence with the mouse. The sequence appears in the textfield at the top left of the UI. The "Replace by" textfield offers you the suggestion for the entity you have selected. The entity type can be selected from the list of types on the right panel containing a series of radio buttons. Here we have selected the string "briefing item" of entity type "OTHER".
In the following figure, you can see the screen that appears after you have clicked the "OK" button at the top of the UI. The blue tag signifies that it is a temporary tag and has been stored in the database. The "Mark Email as Done" button saves the anonymized email. However, if the email contains one or more blue tags, this button does not work, and you have to click on "Mark Done, but Temporarily". The marks the email to be retagged using new permanent tags later. "Re-Examine Later" does not save the email. You can click on "Skip" if you don't want your email in the anonymized dataset. "Show Original" brings up the original email for view. "Retag Email" retags the email from start again.
The following figure shows the presence of green tags. These are tags that are very common but suffer from ambiguity. For example, the word UI in our domain corresponds to some object that may need to be anonymized at times, and may not be anonymized at others. For such terms, we use a green highlight so that they are visible to the anonymizer immediately and he doesn's miss them.
Unmasking the green tag:
