Quasi-randomness of graph balanced cut properties

Abstract

Quasi-random graphs can be informally described as graphs whose edge distribution closely resembles that of a truly random graph of the same edge density. Recently, Shapira and Yuster proved the following result on quasi-randomness of graphs. Let $k \ge 2$ be a fixed integer, $α_1,...,α_k$ be positive reals satisfying $\sum_{i} α_i = 1$ and $(α_1,..., α_k) \neq (1/k,...,1/k)$, and $G$ be a graph on $n$ vertices. If for every partition of the vertices of $G$ into sets $V_1,..., V_k$ of size $α_1 n,..., α_k n$, the number of complete graphs on $k$ vertices which have exactly one vertex in each of these sets is similar to what we would expect in a random graph, then the graph is quasi-random. However, the method of quasi-random hypergraphs they used did not provide enough information to resolve the case $(1/k,..., 1/k)$ for graphs. In their work, Shapira and Yuster asked whether this case also forces the graph to be quasi-random. Janson also posed the same question in his study of quasi-randomness under the framework of graph limits. In this paper, we positively answer their question.