SimpleNLG for German
As of October 2019, a new SimpleNLG adaption for German has been released. It is based on the newer SimpleNLG v4 architecture and comes with a more permissive license. For all practical purposes, you probably want to use this newer adaption instead.
In 2010–2011, I worked on an adaption of the SimpleNLG library for German as part of my studies for my Master’s degree. Since then, I have no longer been working in the field of natural language generation. Unfortunately, this means that my adaption has never quite been finished and “polished up” for release; I still decided to provide what I have for other people who might potentially be interested in it.
If you use SimpleNLG for German in your own work, please cite the following publication:
- Marcel Bollmann. 2011. Adapting SimpleNLG to German. In Proceedings of the 13th European Workshop on Natural Language Generation (ENLG 2011), pages 133–138, Nancy, France.
SimpleNLG is a Java library for generation of grammatically correct sentences. Originally developed for English, I adapted it for German. For more information, please refer to the SimpleNLG tutorial.
Grammatical features covered by this adaption include, among others:
- free ordering of sentence constituents, to account for the more flexible word order of German;
- clusters of modal verbs; and
- support for relative clauses and relative clause extraposition.
As I’m not able to devote time to the development of SimpleNLG for German anymore, this release still has some considerable limitations. The most important ones are:
- it is based on the older version 3 architecture of SimpleNLG, which is now obsolete and also has a more restrictive license;
- there is no proper lexicon—only a “toy lexicon” of around 100 entries is available—and lexicon lookup is quite slow and inefficient; and
- it hasn’t been thoroughly tested for bugs.
If you have any questions specific to my adaption of SimpleNLG for German, please feel free to contact me. For questions about the original SimpleNLG, please refer to the SimpleNLG discussion list instead.