In this module we provide different methods for generating random walks given a graph. The algorithms in mOWL are a variation from the original ones. Graphs obtained from ontologies always have labeled edges, therefore the edge labels are included in the random walks.


Random walks with size \(n\) will include \(n\) nodes with its edges (except in the last node). Therefore a random walk with size \(n\) will be at most \(2n-1\) long.


walker_factory(method_name, num_walks, ...)


DeepWalk(num_walks, walk_length[, alpha, ...])

Implementation of DeepWalk based on [perozzi2014].

Node2Vec(num_walks, walk_length[, p, q, ...])

Implementation of DeepWalk based on [grover2016].

WalkingModel(num_walks, walk_length, outfile)

Base class for walking methods.

Class Inheritance Diagram

Inheritance diagram of mowl.walking.deepwalk.model.DeepWalk, mowl.walking.node2vec.model.Node2Vec, mowl.walking.walking.WalkingModel