This page intends to give an overview of software relevant for people interested in applying ILP (or constrained optimization in general) to problems in Natural Language Processing.

ILP Solvers

Free / Open Source solvers

  • lp_solve: branch and bound, built-in LP solver
  • Coin-OR Cbc: branch-and-cut, can be used with CLP (LP solver)
  • GLPK: branch-and-cut, provides several LP methods
  • SCIP: branch-cut-and price, very fast

Commercial solvers

Higher Order Learning and Inference Frameworks

The following programs/libraries allow users to define ILPs compactly through a higher order representation. Instead of having to write wrapper code, the user can make use of quantification to declaratively generate large sets of linear constraints. In contrast to the higher order modeling frameworks for mathematical programming (such as AMPL) the following frameworks also provide means to learn the weights of soft constraints from data.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License