Cohomological field theories (CohFTs) were defined in the mid 1990s by Kontsevich and Manin to capture the formal properties of the virtual fundamental class in Gromov-Witten theory. A beautiful classification result for semisimple CohFTs (via the action of the Givental group) was proven by Teleman in 2012. The Givental-Teleman classification can be used to explicitly calculate the full CohFT in many interesting cases not approachable by earlier methods.My goal here is to present an introduction to these ideas together with a survey of the calculations of the CohFTs obtained from • Witten's classes on the moduli spaces of r-spin curves, • Chern characters of the Verlinde bundles on the moduli of curves, • Gromov-Witten classes of Hilbert schemes of points of C 2 . The subject is full of basic open questions.The boundary 1 of the Deligne-Mumford compactification is the closed locus parameterizing curves with a least one node, ∂M g,n = M g,n \ M g,n .By identifying the last two markings of a single (n + 2)-pointed curve of genus g − 1, we obtain a morphismSimilarly, by identifying the last markings of separate pointed curves, we obtainwhere n = n 1 + n 2 and g = g 1 + g 2 . The images of both q and r lie in the boundary ∂M g,n ⊂ M g,n .The cohomology and Chow groups of the moduli space of curves are H * (M g,n , Q) and A * (M g,n , Q) .While there has been considerable progress in recent years, many basic questions about the cohomology and algebraic cycle theory remain open. 2
Gromov-Witten classesLet X be a nonsingular projective variety over C, and let