Zubr

A toolkit for building semantic parsing models, especially models for doing semantic parsing in technical domains (e.g., text2code translation)

Zubr

Type
Tool
Author
Kyle Richardson
Description

Zubr is a toolkit for building semantic parsing models, especially models for doing semantic parsing in technical domains (e.g., text2code translation). It also has features for building code retrieval systems. It is written in Python/Cython. 

The system is licensed under the GNU General Public License (GPL). For questions contact Kyle Richardson (firstname@ims.uni-stuttgart.de)

Reference
  • Kyle Richardson and Jonas Kuhn. Learning Semantic Correspondences in Technical Documentation. Proceedings of ACL 2017. [pdf
  • Kyle Richardson and Jonas Kuhn. Function Assistant: A Tool for NL Querying of APIs Proceedings of EMNLP 2017. [pdf] 
  • Kyle Richardson, Jonathan Berant,  Jonas Kuhn. Polyglot Semantic Parsing in APIs Proceedings of NAACL 2018. [pdf] 
Download

The system is hoted here (on Github), and the underlying datasets can be found here

See README.md for installation instructions and the experiments/ directory for more details.  

Jonas Kuhn
Prof. Dr.

Jonas Kuhn

Chair of Foundations of Computational Linguistics, Managing Director of the IMS

To the top of the page