Thanks for downloading our code. This package contains an end-to-end, data-driven concept-to-text generator using PCFGs to learn Document Planning, Content Selection and Surface Realization jointly. If you are going to use/extend this package in your work (and you are most welcome to do so), please cite:
Ioannis Konstas and Mirella Lapata. 2013. A Global Model for Concept-to-Text Generation, JAIR, Volume 48, pages 305-346.
Ioannis Konstas and Mirella Lapata. 2013. Inducing Document Plans for Concept-to-Text Generation. EMNLP, pages 1503-1514, Seattle, Washington, USA.