Satisfiability of boolean formulae (SAT) has been a topic of research in logic and computer science for a long time. In this paper we are interested in understanding the structure of satisfiable and unsatisfiable sentences. In previous work we initiated a new approach to SAT by formulating a mapping from propositional logic sentences to graphs, allowing us to find structural obstructions to 2SAT (clauses with exactly 2 literals) in terms of graphs. Here we generalize these ideas to multi-hypergraphs in which the edges can have more than 2 vertices and can have multiplicity. This is needed for understanding the structure of SAT for sentences made of clauses with 3 or more literals (3SAT), which is a building block of NP-completeness theory. We introduce a decision problem that we call GraphSAT, as a first step towards a structural view of SAT. Each propositional logic sentence can be mapped to a multi-hypergraph by associating each variable with a vertex (ignoring the negations) and each clause with a hyperedge. Such a graph then becomes a representative of a collection of possible sentences and we can then formulate the notion of satisfiability of such a graph. With this coarse representation of classes of sentences one can then investigate structural obstructions to SAT. To make the problem tractable, we prove a local graph rewriting theorem which allows us to simplify the neighborhood of a vertex without knowing the rest of the graph. We use this to deduce several reduction rules, allowing us to modify a graph without changing its satisfiability status which can then be used in a program to simplify graphs. We study a subclass of 3SAT by examining sentences living on triangulations of surfaces and show that for any compact surface there exists a triangulation that can support unsatisfiable sentences, giving specific examples of such triangulations for various surfaces.
The behavior of a certain random growth process is analyzed on arbitrary regular and non-regular graphs. Our argument is based on the Expander Mixing Lemma, which entails that the results are strongest for Ramanujan graphs, which asymptotically maximize the spectral gap. Further, we consider ErdH{o}s--Renyi random graphs and compare our theoretical results with computational experiments on flip graphs of point configurations. The latter is relevant for enumerating triangulations.
In this note, we show how to obtain a characteristic power series of graphons -- infinite limits of dense graphs -- as the limit of normalized reciprocal characteristic polynomials. This leads to a new characterization of graph quasi-randomness and another perspective on spectral theory for graphons, a complete description of the function in terms of the spectrum of the graphon as a self-adjoint kernel operator. Interestingly, while we apply a standard regularization to classical determinants, it is unclear how necessary this is.
More than ten years ago in 2008, a new kind of graph coloring appeared in graph theory, which is the {it rainbow connection coloring} of graphs, and then followed by some other new concepts of graph colorings, such as {it proper connection coloring, monochromatic connection coloring, and conflict-free connection coloring} of graphs. In about ten years of our consistent study, we found that these new concepts of graph colorings are actually quite different from the classic graph colorings. These {it colored connection colorings} of graphs are brand-new colorings and they need to take care of global structural properties (for example, connectivity) of a graph under the colorings; while the traditional colorings of graphs are colorings under which only local structural properties (adjacent vertices or edges) of a graph are taken care of. Both classic colorings and the new colored connection colorings can produce the so-called chromatic numbers. We call the colored connection numbers the {it global chromatic numbers}, and the classic or traditional chromatic numbers the {it local chromatic numbers}. This paper intends to clarify the difference between the colored connection colorings and the traditional colorings, and finally to propose the new concepts of global colorings under which global structural properties of the colored graph are kept, and the global chromatic numbers.
Alon and Yuster proved that the number of orientations of any $n$-vertex graph in which every $K_3$ is transitively oriented is at most $2^{lfloor n^2/4rfloor}$ for $n geq 10^4$ and conjectured that the precise lower bound on $n$ should be $n geq 8$. We confirm their conjecture and, additionally, characterize the extremal families by showing that the balanced complete bipartite graph with $n$ vertices is the only $n$-vertex graph for which there are exactly $2^{lfloor n^2/4rfloor}$ such orientations.