RockIt: Exploiting Parallelism and Symmetry for MAP Inference in Statistical Relational Models
This work addresses efficiency and quality challenges in MAP inference for statistical relational models, offering a domain-specific improvement over existing systems.
The paper tackles the problem of MAP inference in statistical relational models by introducing RockIt, a query engine that uses cutting plane aggregation and parallelization to improve efficiency and result quality, outperforming state-of-the-art systems like Alchemy, Markov TheBeast, and Tuffy in benchmarks.
RockIt is a maximum a-posteriori (MAP) query engine for statistical relational models. MAP inference in graphical models is an optimization problem which can be compiled to integer linear programs (ILPs). We describe several advances in translating MAP queries to ILP instances and present the novel meta-algorithm cutting plane aggregation (CPA). CPA exploits local context-specific symmetries and bundles up sets of linear constraints. The resulting counting constraints lead to more compact ILPs and make the symmetry of the ground model more explicit to state-of-the-art ILP solvers. Moreover, RockIt parallelizes most parts of the MAP inference pipeline taking advantage of ubiquitous shared-memory multi-core architectures. We report on extensive experiments with Markov logic network (MLN) benchmarks showing that RockIt outperforms the state-of-the-art systems Alchemy, Markov TheBeast, and Tuffy both in terms of efficiency and quality of results.