The moduli space of 4-dimensional nilpotent complex associative algebras (Q2250781)

The authors' method of constructing moduli spaces of low dimensional complex and real Lie and associative algebras is based on the principle that such algebras are either simple, or can be constructed as extensions of lower dimensional algebras. The authors interpret the classical theory of extensions in the language of codifferentials, and they give a description of the theory of extensions of an algebra \(W\) by an algebra \(M\). Consider the exact sequence \(0\rightarrow M\rightarrow V\rightarrow W\rightarrow 0\). Letting \(\delta\in C^2(W)\) and \(\mu\in C^2(M)\) represent the algebra structure on \(W\) and \(M\) respectively, it is possible to view \(\mu\) and \(\delta\) as elements on \(C^2(V)\). Letting \(T^{k,l}\) denote the subspace of the component \(T^{k+l}(V)\) of the tensor algebra \(T(V)\) defined inductively by \(T^{0,0}=k\), \(T^{k,l}=M\otimes T^{k-1,l}\oplus V\otimes T^{k,l-1}\) and then \(C^{k,l}=\text{Hom}(T^{k,l},M)\subseteq C^{k+l}(V)\), it is possible to write the algebra structure on \(V\) as \(d=\delta+\mu+\lambda+\psi,\) with \(\lambda\in C^{1,1}\) and \(\psi\in C^{0,2}\) (and \(\mu\in C^{2,0}\)). The associativity gives \([\delta,\lambda]+\frac12[\lambda,\lambda]+[\mu,\psi]=0\), the Maurer-Cartan equation, \([\mu,\lambda]=0\), the compability condition, and \([\delta+\lambda,\psi]=0\), the cocycle condition. Defining \(D_\mu\) by \(D_\mu(\phi)=[\mu,\phi]\), \(D_\mu\) is a differential on \(C(V)\). This leads to a definition of of the structures \(d\) and \(d^\prime\) being \textit{equivalent in the restricted sense}, and the cohomology classes of those \(\lambda\) for which a solution of the MC equations exists, determine distinct restricted equivalence classes of extensions. The authors use the classification of extensions of \(\delta\) on \(W\) by \(\mu\) on \(M\) given by (1) Equivalence classes of \(\overline\lambda\in H^{1,1}_\mu\) which satisfy the MC equations \([\delta,\lambda]+\frac12[\lambda,\lambda]+[\mu,\psi]=0\) for some \(\psi\in C^{0,2}\) under the action of a group \(G_{\delta,\mu}\), and (2) Equivalence classes of \(\{\overline\tau\}\in H^{0,2}_{\mu,\delta+\lambda}\) under the action of a group \(G_{\delta,\mu,\lambda}\). The fundamental theorem of finite dimensional algebras allows to restrict the consideration of extensions to two central cases: Firstly, consider the extensions where \(\delta\) is a semisimple algebra structure on \(W\) and \(\mu\) is a nilpotent algebra structure on \(M\). Then the classification reduces to considering equivalence classes of \(\lambda\). Secondly, extensions of the trivial algebra structure \(\delta=0\) on a \(1\)-dimensional space \(W\) by a nilpotent algebra \(\mu\) can be considered. Then the classification theorem above determines the equivalence classes of extensions. Under some conditions, the classification of extensions up to equivalence is sufficient to classify the algebras. Thus the equivalence classes of the module structure \(\lambda\) determine the algebra up to isomorphism. For nilpotent algebras, this does not hold, and the idea for the construction of nilpotent algebras is given in the present article. It has a basis in the fact that any nilpotent algebra determines a unique kernel and a \textit{core}, and the algebra is given by an extension of its core by the kernel in a particularly simple manner. The classification of complex bilinear forms is an important component of the construction. This means that the authors give a classification of complex bilinear forms on a \(2\)-dimensional complex vector space and then use this classification to give a construction of \(3\)-dimensional complex algebras that agrees with earlier results. Then this is generalized one step to give the construction of nilpotent \(4\)-dimensional complex algebras. It should be mentioned that the article gives a decomposition of the moduli space into strata which are are given by deformation theory. The construction given in the article interpret the strata as projective orbifolds which needs to be identified in some natural way. This made the classification possible. The article in some sense lists the different type of algebras, and uses the last section for the analysis of the results. It is very nice and contains important results which have been awaited as a natural part in the series of classification of algebras.