UDepLambda is a framework to convert Universal Dependencies trees to Logical Forms. It maps natural language to logical forms in an almost language-independent framework. For more details, please refer to our papers below.
- Java 8
- Ant 1.8.2 or higher
Run the following commands in a terminal:
git clone https://github.com/sivareddyg/UDepLambda.git
cd UDepLambda
git submodule update --init --recursive lib
git submodule update --init --recursive lib_data/ud-models-v1.3
git submodule update --init --recursive lib_data/ud-models-v1.2
ant build
File -> Import -> Existing Projects into Workspace -> Select root directory as UDepLambda -> Finish
For English:
cat input-english.txt | sh run-english.sh
For German:
cat input-german.txt | sh run-german.sh
For Spanish:
cat input-spanish.txt | sh run-spanish.sh
Check debug.txt to see step by step derivation.
- See FreebaseExperiments.md for Freebase semantic parsing experiments.
- See SimplifiedLogicForm.md for output formats.
- See doc for additional documentation.
We welcome any kind of contributions. Please see CONTRIBUTING.md for instructions on how to contribute.
UDepLambda is distributed as Apache 2.0 license. However, depending on the pipeline and resources you use, you may have to get additional licenses. For example, the default pipeline uses the Stanford CoreNLP Pipeline, and the logical expression parsing engine is based on Cornell Semantic Parsing Framework. You can see the list of libraries we use at https://bitbucket.org/sivareddyg/udeplambda-lib/src.
@article{reddy2017universal,
title={Universal Semantic Parsing},
author={Reddy, Siva and T{\"a}ckstr{\"o}m, Oscar and Petrov, Slav and Steedman, Mark and Lapata, Mirella},
journal={arXiv preprint arXiv:1702.03196},
year={2017}
url = {https://arxiv.org/pdf/1702.03196.pdf}
}
Universal Semantic Parsing
Siva Reddy, Oscar Täckström, Slav Petrov, Mark Steedman, Mirella Lapata
arXiv:1702.03196, 2017.
Universal Dependencies to Logical Forms with Negation Scope
Federico Fancellu, Siva Reddy, Adam Lopez, Bonnie Webber
arXiv:1702.03305, 2017.