HETA

Intro

Hierarchical Edge Type Analysis. An algorithm for detecting edge types using common neighbor concept.

This is a rewritten of: https://github.com/canslab1/Identification-Algorithm/ , into convenient function that take NetworkX Graph objects, and to be compatible with Python3.

Install

git clone https://github.com/wcchin/HETA.git
cd HETA
pip install .
# or 
# pip install -e .

Usage

Please check the test.py script.

In simple:

### import necessary packages and heta package
import networkx as nx
import heta

### read file as nx.Graph() object
fp = 'data/net/14p.net'
g = nx.Graph(nx.read_pajek(fp))

### run algorithm using default parameter
g, ext_dic, int_dic = heta.bridge_or_bond(g)

### get all edge types
for u, v, d in g.edges(data=True):
    print(u, v, d['type'])

### draw the results
import matplotlib.pyplot as plt
heta.draw_result(g, layout='spring')
plt.show()

### external and internal threshold
print(ext_dic, int_dic)

### fingerprint analysis
### proportion of ['bond', 'local', 'global', 'silk']
print(heta.fingerprint(g))

Article reference

Beyond Bond Links in Complex Networks: Local Bridges, Global Bridges and Silk Links

Highlights

• An identification algorithm for determining hierarchical link types is described.
• Multi-hierarchy level link types are identified using the common neighbor concept.
• Two applications are demonstrated: fingerprint analysis and network partitioning.
• Fingerprint analysis is used to compare topological structures among networks.
• Hierarchical link structures are used to partition networks into communities.

Abstract

Many network researchers use intuitive or basic definitions when discussing the importance of strong and weak links and their roles. Others use an approach best described as “if not strong, then weak” to determine the strengths and weaknesses of individual links, thus deemphasizing hierarchical network structures that allow links to express different strength levels. Here we describe our proposal for a hierarchical edge type analysis (HETA) algorithm for determining link types at multiple network hierarchy levels based on the common neighbor concept plus statistical factors such as bond links, kth-layer local bridges, global bridges, and silk links—all generated during long-term network development and evolution processes. Two sets of networks were used to validate our proposed algorithm, one consisting of 16 networks employed in multiple past studies, and one consisting of two types of one-dimensional small-world networks expressing different random rewiring or shortcut addition probabilities. Two applications with potential for developmental contributions are demonstrated: a network fingerprint analysis framework, and a hierarchical network community partition method.

Keywords: Network topology, Hierarchy of links, Common neighbor concept, Fingerprint analysis, Hierarchical community partition, Edge type analysis

article url: https://www.sciencedirect.com/science/article/pii/S0378437119306375

cite this as:
Huang C. Y., Chin, W. C. B., Fu, Y. H., & Tsai, Y. S. (in press) Beyond bond links in complex networks: Local bridges, global bridges and silk links. Physica A: Statistical Mechanics and its Applications. DOI: https://doi.org/10.1016/j.physa.2019.04.263 .