What is the effect of sparse data on SVM performance?

SVM performs very well on sparse data like text vectors.

Sparse datasets contain many zero-valued features. SVM handles this efficiently because only support vectors matter in defining decision boundaries. Linear SVMs are especially efficient in sparse high-dimensional spaces.