The Artificial Benchmark for Community Detection with Outliers and Overlapping Communities ( ABCD+o 2 )

Jordan Barrett, Ryan DeWolfe, Bogumił Kamiński, Paweł Prałat, Aaron Smith, François Théberge · 2025

Read The Artificial Benchmark for Community Detection with Outliers and Overlapping Communities ( ABCD+o 2 ) by Jordan Barrett, Ryan DeWolfe, Bogumił Kamiński, Paweł Prałat, Aaron Smith, François Théberge (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-05486.txt.

Description

Jordan Barrett and co-authors present a 2025 paper on fake networks for testing community finders. Real networks hide their true communities, so tests need models with known groups. The ABCD model builds random graphs with community structure. Degree and community sizes follow power-law patterns, like real networks. It runs faster than the classic LFR model, stays easier to read, and allows exact analysis.

A past variant added outliers, nodes that belong to no group. The new variant also allows overlapping communities, where a node joins several groups at once. Overlap is common in real social and work networks. Finder tools can now be graded on graphs that mix groups, outliers, and overlap in known amounts. Honest grading matters: a finder that shines on clean graphs may fail on the messy overlap of real data.

Catalog record

CREATORJordan Barrett, Ryan DeWolfe, Bogumił Kamiński, Paweł Prałat, Aaron Smith, François Théberge
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-05486

Browse the full library · About the Tower of Babel