Current and emerging networks have a duty to promote interconnections for new collaborative research and business between academia and industry. Multi-layered networks require descriptions of relations among nodes from different network nodes to enable diverse organizations to collaborate in research activities such as data mining. But how should various networks share data in the presence of noise, partial access, and other barriers to network connections? Chen et al. present an algorithm for discovering interference among communicating nodes in multi-layered networks.

The authors argue: (1) most prevailing multi-layered network research assumes the existence of both within-layer connections and cross-layer dependencies--they do not reflect real-world network applications that are susceptible to noise and restricted access; and (2) collaborative filtering methods can be used to deduce the missing cross-layer dependencies in multi-layered networks and to capture the implicit dependencies between network users and items. Undeniably, accurate cross-layer dependency intrusion results in any multifaceted network should examine the dependencies as well as within-layer and cross-layer connections among network layers. Consequently, they frame the cross-layer dependence meddling network issues as a normalized optimization problem. The authors present an algorithm for probing the cross-layer reliance on multifaceted networks; examine its gilt-edged, convergent quality and complicatedness; and outline its alternatives and generalities.

Experiments performed on real-life datasets evaluate the effectiveness and scalability of the proposed algorithm. The cross-layer interference problems are investigated using the within-layer similarities of chemical, drug, disease, and gene (protein-protein-interaction) networks; the chemical-drug and disease-protein dependencies; and drug-disease interactions. The experimental results reveal that the new algorithm outperforms existing competitive algorithms for decision making about cross-layer dependency in multifaceted networks. Clearly, the authors offer unique views for promoting communications among different related networks, and present practical and effective algorithms for looking into similar information among different networks. All network administrators and designers of diverse network applications ought to read the paper’s insightful ideas for coping with existing and future barriers to internetwork communication issues.