Chinese Whispers for Python¶

This is an implementation of the Chinese Whispers graph clustering algorithm in Python.

  • https://pypi.org/project/chinese-whispers/
  • https://chinese-whispers.readthedocs.io/
  • https://github.com/nlpub/chinese-whispers

Version Information¶

In [1]:
from matplotlib import __version__ as plt_version
from networkx import __version__ as nx_version

from chinese_whispers import __version__ as cw_version

print(f"Chinese Whispers {cw_version}")
print(f"NetworkX {nx_version}")
print(f"matplotlib {plt_version}")

Chinese Whispers 0.9.0
NetworkX 3.4.2
matplotlib 3.10.1

Clustering¶

In [2]:
import networkx as nx

from chinese_whispers import aggregate_clusters, chinese_whispers
In [3]:
G = nx.karate_club_graph()
In [4]:
# Perform clustering of G, parameters weighting and seed can be omitted
chinese_whispers(G, weighting="top", seed=1337)

# Print the clusters in the descending order of size
print("ID\tCluster\n")

for label, cluster in sorted(aggregate_clusters(G).items(), key=lambda e: len(e[1]), reverse=True):
    print(f"{label}\t{cluster}\n")

ID	Cluster

24	{32, 33, 8, 9, 14, 15, 18, 20, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31}

3	{0, 1, 2, 3, 4, 7, 10, 11, 12, 13, 17, 19, 21}

7	{16, 5, 6}

Visualization¶

In [5]:
%config InlineBackend.figure_formats = ['svg']

import matplotlib.pyplot as plt
In [6]:
colors = [1. / G.nodes[node]["label"] for node in G.nodes()]

nx.draw_networkx(G, cmap=plt.get_cmap("jet"), node_color=colors, font_color="white")
No description has been provided for this image

Citation¶

  • Ustalov, D., Panchenko, A., Biemann, C., Ponzetto, S.P.: Watset: Local-Global Graph Clustering with Applications in Sense and Frame Induction. Computational Linguistics 45(3), 423–479 (2019)
@article{Ustalov:19:cl,
  author    = {Ustalov, Dmitry and Panchenko, Alexander and Biemann, Chris and Ponzetto, Simone Paolo},
  title     = {{Watset: Local-Global Graph Clustering with Applications in Sense and Frame Induction}},
  journal   = {Computational Linguistics},
  year      = {2019},
  volume    = {45},
  number    = {3},
  pages     = {423--479},
  doi       = {10.1162/COLI_a_00354},
  publisher = {MIT Press},
  issn      = {0891-2017},
  language  = {english},
}

In case you require higher performance, please consider our Java implementation that also includes other graph clustering algorithms: https://github.com/nlpub/watset-java.

In [ ]: