Graph Isomorphism

Graph Isomorphism
Name	Graph Isomorphism
Field	Mathematics; Computer science
Related	Graph theory; Isomorphism (mathematics); Computational complexity theory

Graph Isomorphism is the relation between two finite undirected graphs that can be mapped onto each other by a bijection of vertices preserving adjacency, forming a central problem in graph theory and theoretical computer science. It connects to structural classification problems studied at Princeton University, Massachusetts Institute of Technology, and Bell Labs and has implications for complexity classes explored by researchers at Institute for Advanced Study and Microsoft Research. The problem interacts with combinatorial group theory studied at Cambridge University and algorithmic graph theory developed at Stanford University.

Definitions

Two graphs G and H are said to be isomorphic if there exists a bijection f between their vertex sets such that any vertices u and v are adjacent in G exactly when f(u) and f(v) are adjacent in H. This bijection, called an isomorphism, induces an automorphism when G = H, a concept analyzed by researchers at École Normale Supérieure and Harvard University. The set of all automorphisms of a graph forms a group studied within group theory and connected to actions of symmetric group on vertex labels. Related notions include labeled graphs used in work at University of California, Berkeley and unlabeled graphs considered in classical texts from Oxford University Press.

Computational Complexity

The decision problem asking whether two input graphs are isomorphic lies in NP but is not known to be NP-complete under standard assumptions, a status debated at Clay Mathematics Institute and in workshops at SIAM conferences. It occupies a unique place among problems like Integer Factorization and Graph Coloring discussed at ETH Zurich and has inspired complexity class discussions involving co-AM and reductions studied at Cornell University. The problem is in quasi-polynomial time following breakthroughs celebrated at International Congress of Mathematicians sessions and is related to parameterized complexity analyses at Carnegie Mellon University.

Algorithms and Methods

Classical algorithms include the Weisfeiler–Lehman test developed at Steklov Institute of Mathematics and refined by researchers at Max Planck Institute; color refinement techniques from University of Waterloo; and individualization-refinement frameworks employed by software from Wolfram Research and teams at University of Vienna. Group-theoretic methods pioneered by work at University of Chicago and Princeton University exploit the structure of permutation groups such as the alternating group and symmetric group to test equivalence. Recent algorithmic advances yielding quasi-polynomial-time procedures were achieved by researchers affiliated with University of California, San Diego and presented in venues like ACM Symposium on Theory of Computing and European Symposium on Algorithms.

Special Graph Classes and Cases

For many restricted classes the problem is tractable: trees were settled by methods from Bell Labs and textbooks from Cambridge University Press; planar graphs admit linear-time tests based on results originating at University of Tokyo and ENSTA Paris; bounded-degree graphs and bounded-eigenvalue multiplicity cases were addressed by teams at University of Oxford and Rothamsted Research. Interval graphs, chordal graphs, and comparability graphs are solvable using structural decomposition techniques developed at Yale University and University of Toronto. Cayley graphs and strongly regular graphs link the problem to algebraic combinatorics studied at Hebrew University of Jerusalem and University of Edinburgh.

Practical Applications

Graph isomorphism algorithms are used in chemical informatics for molecular structure matching in industry labs at Pfizer and Merck and in cheminformatics tools developed at European Bioinformatics Institute. Isomorphism testing underpins pattern recognition in computer vision research at Carnegie Mellon University and MIT Computer Science and Artificial Intelligence Laboratory, matching substructures in social network analysis conducted at Facebook and Google Research. It supports database schema matching explored by teams at Oracle Corporation and IBM Research, and code clone detection in software engineering studied at Microsoft Research and ETH Zurich.

Historical Development and Key Results

Early formalizations trace to foundational work in Euler-era graph studies and 20th-century developments at University of Göttingen and University of Cambridge. Significant milestones include the introduction of automorphism group techniques linked to Emmy Noether's influence on algebra, the Weisfeiler–Lehman hierarchy studied at Moscow State University, and the classification of tractable special cases by researchers at Columbia University and Rutgers University. Landmark algorithmic progress in the 21st century, yielding quasi-polynomial-time algorithms, was announced in venues such as International Colloquium on Automata, Languages and Programming and validated in presentations at NeurIPS and FOCS. Ongoing open questions continue to motivate research groups at Princeton University, University of Cambridge, and University of California, Berkeley.

Category:Graph theory