In an earlier work R. A. Duke and V. Rödl, The Erdős–Ko–Rado theorem for small families, J Combin Theory Ser A 65(2) (1994), 246–251 it was shown that for t a fixed positive integer and κ a real constant, 0 < κ < 1/2, if n is sufficiently large each family 𝒜 of ⌊κ n ⌋‐element subsets of [ n ] of size N (linear in n ) contains a t ‐intersecting subfamily of size at least (1 − o (1))κ N . Here we consider the case when t , the intersection size, is no longer bounded, specifically t = ⌊τ n ⌋ for 0 < τ < κ. We show that for sufficiently large n and N each family of this type contains an r ‐wise t ‐intersecting subfamily of size at least N 1−δ , and that, apart for the size of δ, this result is the best possible.
No takes yet. Share an insight, caveat, or question.
Duke et al. (2003) studied this question.
Synapse has enriched one closely related paper. Consider it for comparative context: