At the start of the algorithm, a (generally small) subset of the data points have labels (or classifications).
In comparison with other algorithms[3] label propagation has advantages in its running time and amount of a priori information needed about the network structure (no parameter is required to be known beforehand).
Membership in a community changes based on the labels that the neighboring nodes possess.
When many such dense (consensus) groups are created throughout the network, they continue to expand outwards until it is impossible to do so.
Text classification utilizes a graph-based technique, where the nearest neighbor graph is built from network embeddings, and labels are extended based on cosine similarity by merging these pseudo-labeled data points into supervised learning.
[4] In contrast with other algorithms label propagation can result in various community structures from the same initial condition.