Abstract
Node embedding has recently shown state-of-the-art performance in various network analysis tasks. However,most of the existing node embedding methods do not considerthe uncertainty of the input data, which is often the case inpractice. This work offers an empirical evaluation of the typicalnode embedding methods when applied on uncertain networks.Precisely, we examine the performance of embedding vectorsobtained by these methods in a set of downstream tasks. Tothis end, we employ a wide range of uncertain networks andtraditional prepossessing techniques for dealing with uncertainty.Our findings suggest that the existing node embedding methodsperform practically well on networks with uncertainty once thenetwork data is appropriately prepossessed.
Users
Please
log in to take part in the discussion (add own reviews or comments).