NoNCrossRank

This is an implementation of NoNCrossRank algorithm with Python.

The goal of this algorithm is to solve:
Given: (1) an Network of Networks (NoN) R = <G, A, θ>, and (2) the query vectors e_i (i = 1, ..., g);
Find: ranking vectors r_i for the nodes in the domain-specific networks A_i (i = 1, ..., g).

Functions

• __init__.py: program entry;
• CR_CrossValidation.py: cross rank cross validation;
• CR_Precomputation.py: precompute the Anorm and Ynorm;
• CR.py: ranking process;
• J_CR.py: compute objective fuction;
• AUCEvaluation.py AUC evaluation process;
• AUCValue.py: compute AUC value;

Input/Output Format

- Input

G: the adjacency matrix of main network
A: domain specific networks A = (A₁, ..., A_g)
θ: the one-to-one mapping function (mapping main node to domain-specific network)
e_i: the query vector for A_i (i = 1, ..., g)

- Output

r_i: the ranking vector for A_i (i = 1, ..., g)
AUCValue: the AUC value is used for evaluation

Reference

Ni, J., Tong, H., Fan, W., & Zhang, X. (2014, August). Inside the atoms: ranking on a network of networks. In Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 1356-1365). ACM.