Is there a method to measure the similarity between undirected graph vertices?

Question

I'm doing some research on User Identity Resolution. Assume i can get two undirected graphs of a person, one is the friendship in Twitter of that person, the other is the friendship in Facebook of that person.(For Twitter, i remove the single relation and preserve the mutual relation.).

I want to calculate the similarity between vertices in that undirected graphs to get the possible pairs seem to be same persons.

Let G_T(n_T*n_T) and G_F(n_F*n_F) be two undirected graph. And the result may be a n_F*n_T matrix, the each value in the result matrix represent the similarity between vertices in above graphs.

And i found Measure of Similarity between Graph Vertices: Applications to Synonym Extraction and Web Searching. When i apply the method that thesis proposed to my data, the result isn't satisfied.

Is there other method to do that? Or other method to process the data?