Berge conjecture

Berge conjecture
Name	Berge conjecture
Field	Graph theory, Combinatorics
Inventor	Claude Berge
Introduced	1960s
Status	open (partial results)

Berge conjecture is a conjecture in Graph theory and Combinatorics proposed by Claude Berge concerning matchings and covers in general graphs. It connects structural properties of finite graphs with extremal bounds influenced by classical results such as Konig's theorem, Tutte's theorem, and the Erdős–Gallai theorem. The conjecture has stimulated work involving researchers affiliated with institutions like the Institut Henri Poincaré, Princeton University, Massachusetts Institute of Technology, and journals such as Annals of Mathematics and Journal of Combinatorial Theory.

Statement of the conjecture

The conjecture asserts that for every finite graph G there exists a collection of disjoint matchings whose union covers all edges of G subject to constraints relating the size of a minimum edge cover and the size of a maximum matching. It proposes an equality or inequality linking the edge-cover number, matching number, and parameters that generalize classical identities found in bipartite settings like Konig's theorem and the characterization of perfect matchings in Tutte's theorem. The formulation inspired variations involving parameters from extremal graph theory, perfect matching structure, and decomposition results pursued by authors at University of Cambridge, University of California, Berkeley, and École Normale Supérieure.

Historical background and motivation

The conjecture originates in the mid-20th century work of Claude Berge on matching theory and combinatorial optimization, developed in dialogue with contemporaries at University of Paris and researchers such as Paul Erdős and Dénes Kőnig. It was motivated by attempts to extend Konig's theorem from bipartite graphs to nonbipartite contexts and to unify results like Tutte's theorem and the Gallai–Edmonds decomposition under a common combinatorial principle. Early motivation also came from problems in network flow and applications in scheduling and design studied at Bell Labs and the IBM Research labs, which spurred investigations by mathematicians connected to Princeton University and Stanford University.

Known results and partial proofs

Partial confirmations of the conjecture were obtained in special classes of graphs: for bipartite graphs the statement reduces to Kőnig's theorem proved by Dénes Kőnig; for planar graphs and graphs embeddable on fixed surfaces researchers from University of Waterloo and TU Berlin established bounds relying on the Four Color Theorem and structural decompositions used by teams at University of Oxford. Results for graphs with high minimum degree, claw-free graphs studied by groups at Carnegie Mellon University, and bridgeless cubic graphs examined in work connected to Cambridge University provided further validations under additional hypotheses. Techniques employed include the Gallai–Edmonds decomposition, polyhedral methods related to the matching polytope developed by collaborators at INRIA and the Courant Institute, and algorithmic proofs drawing on augmenting path methods from Edmonds' blossom algorithm and advances at Bell Labs and AT&T Bell Laboratories.

Counterexamples and related problems

No definitive counterexample to the original conjecture is known, but several constructions demonstrate the necessity of added hypotheses; these constructions were developed by researchers at University of British Columbia, Institute for Advanced Study, and Princeton University and relate to pathological instances studied in extremal graph theory and design theory. Closely related problems include the Berge–Fulkerson conjecture, the 1-factorization conjecture for regular graphs explored at ETH Zurich and Imperial College London, and packing and covering dualities investigated by scholars affiliated with CNRS and Max Planck Institute for Mathematics. Work linking the conjecture to counterexamples in generalized hypergraph settings involved collaborations between teams at University of Illinois Urbana-Champaign and Tel Aviv University.

Applications and implications

If resolved, the conjecture would impact algorithmic matching theory and optimization results used in operations research groups at MIT and INSEAD, improve bounds in network design problems studied at Bell Labs and the IBM Research division, and influence structural graph theory work at Princeton University and ETH Zurich. Consequences would include tighter integrality gap estimates for the matching polytope used in approximation algorithms developed at Stanford University and clarity on decomposition theorems underpinning results in chemical graph theory pursued at University of Cambridge and University of Tokyo.

Variants and generalizations

Multiple variants generalize the conjecture to directed graphs, hypergraphs, and weighted settings; these have been proposed by researchers at University of Pennsylvania, University of Michigan, Seoul National University, and Australian National University. Notable generalizations include hypergraph matching analogues related to the Ryser conjecture and fractional relaxations connecting to the LP duality framework used in combinatorial optimization studies at Columbia University and University of Chicago. Further extensions explore connections with the Erdős–Stone theorem and structural decompositions akin to the Szemerédi regularity lemma developed by groups at Rutgers University and Microsoft Research.

Category:Conjectures in graph theory