Dockerfile for the Joty 2013 RST Parser

About the Dockerfile

This Dockerfile provides a full environment for running the Joty 2013 RST Parser on bare text. It installs all necessary dependencies. It additionally contains a simple patch to re-train the discourse segmenter via commandline and for convenience, a bash script to robustly parse a larger collection of texts without having issues of old temporary files.

Note: Do not run the same parser instance in parallel, as the parser relies on hardcoded temporary files.

Usage

docker build -t joty-2013-rstparser .
docker run -v /Path/to/text/files:/samples -i -t joty-2013-rstparser
cd Discourse_Parser_Dist/
./parse.sh /samples/example.txt
cat /samples/example.txt.rst

Note: Building the dockerimage can take a while, since this project relies on old versions of numpy and sklearn, which need to compile extensions from source.

About the RST Parser

See http://alt.qcri.org/tools/discourse-parser/

Developer

• Shafiq Joty sjoty@qf.org.qa

References

• Shafiq Joty, Giuseppe Carenini, Raymond Ng and Yashar Mehdad. Combining Intra- and Multi-sentential Rhetorical Parsing for Document-level Discourse Analysis.In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (ACL 2013), Sofia, Bulgaria
• Shafiq Joty, Giuseppe Carenini and Raymond Ng. A Novel Discriminative Framework for Sentence-Level Discourse Analysis. In Proceedings of the Conference on Empirical Methods in Natural Language Processing and the Conference on Natural Language Learning (EMNLP-CoNLL 2012), Jeju, Korea.