Conservative Maltsev Constraint Satisfaction Problems
Manuel Bodirsky, Andrew Moorhead
TL;DR
This work addresses CSPs over finite structures with a conservative Maltsev polymorphism, showing they can be solved in the complexity class $\oplus L$ by a symmetric linear $\mathbb Z_2$-Datalog program. The authors develop a detailed structure theory for conservative minority algebras, including a tree-based representation (via leaf algebras) and a finite invariant-relations basis, then define the nested classes $\mathbf P^{n,k}$ and structures $\mathfrak P^{n,k}$ to build a pp-definable reduction scheme. Central to the approach are the Solve and Reduce procedures, which recursively reduce CSP$(\mathfrak P^{n,k})$ to the base case CSP$(\mathfrak P^{n,1})$ and solve a $\mathbb Z_2$-linear system; the process is implemented in symmetric linear $\mathbb Z_2$-Datalog, placing the problem in $\oplus L$. A companion paper completes the pp-construction from any conservative Maltsev structure to the basic class, reinforcing the universality of the approach. The results yield a full complexity-theoretic classification (up to logspace reductions) for conservative Maltsev CSPs and demonstrate a powerful descriptive-complexity bridge via $G$-Datalog.
Abstract
We show that for every finite structure B with a conservative Maltsev polymorphism, the constraint satisfaction problem for B can be solved by a symmetric linear Z2-Datalog program, and in particular is in the complexity class parity-L. The proof has two steps: we first present the result for a certain subclass whose polymorphism algebras are hereditarily subdirectly irreducible. We then show that every other structure in our class can be primitively positively constructed from one of the structures in the subclass. The second step requires different techniques and will be presented in a companion article.
