There are no reviews yet. Be the first to send feedback to the community and the maintainers!
Repository Details
We design a Monte Carlo based method, UniWalk, to enable the fast top-k SimRank computation over large undirected graphs. UniWalk directly locates the top-k similar vertices for any single source vertex u via approximately R sampling paths originating from u only. Related work has been published in ICDE 2017 ("UniWalk: Unidirectional Random Walk Based Scalable SimRank Computation over Large Graph"). Furthermore, in order to overcome the "rectification issue" in the original UniWalk, we design a path enumeration strategy to gain higher precision, which computes SimRank with the path probabilities instead of path frequencies in sampling. We call it the optimized UniWalk here, which is shown in the "UniWalk/src/simrank/SingleRandomDev.java" file(the basic UniWalk is shown in "UniWalk/src/simrank/SingleRandomWalk.java").
