Motives associated to graphs (Q1000338)

Calculations of Feynman integrals arising in perturbative quantum field theory reveal interesting patterns of zeta and multiple zeta values. These are motivic in origin, what is well known, even if the definition itself of what is a motive is still, as the author recognises, a difficult issue. The purpose of the paper is to give an expository account of some general mathematical aspects of these ``Feynman motives''. In fact, as stated by the author, the paper is a report on recent results on outstanding problems concerning motives associated to graphs. However, there is little in the paper about multiple zeta numbers and Feynman amplitudes (their relationship being as of now still rather unclear). The focus is instead on the Hopf algebra properties of the motivic construction. The paper is based on work of the author with \textit{H. Esnault} and \textit{D. Kreimer} [Commun. Math. Phys. 267, 181--225 (2006; Zbl 1109.81059)]. Some of the subjects dealt with in the paper are: perturbed Gaussian integration, the combinatorial structure of the graph hypersurface, the motive itself, motives and graph homology, and renormalization.