When the data contains high percentage of zero rather than non-zero values, then it is called density sparse data.

High-dimensional sparse features (e.g. from One Hot Vector) relate to the Curse of Dimensionality.