I abandoned in the middle, as the section will probably be re-written

The symbol is still X (and x for the value) but the macro was changed to \obsFeatures, to distinguish from other sets of (possibly unobservable) of features.

We now have a description of our approach.
However, there are certain details that need to be completed

Corrected the caption for intro2problem figure and removed completed to-do

Major change: added image to describe the problem we consider

The figure should explain in an intuitive manner what we do in the paper

We should add the identity of the judge in the dataset

Updated the structure of the introduction, mainly to make the problem more clear