[Mova](https://md.mova.elamur.ai/free/en/) · [this page in HTML](https://mova.elamur.ai/free/en/docs/papers/wahlster-1995-verbmobil-drt-based-translation-spontaneous/)

[home](https://md.mova.elamur.ai/free/en/) / [docs](https://md.mova.elamur.ai/free/en/docs/) / [papers](https://md.mova.elamur.ai/free/en/docs/papers/) / [Machine translation](https://md.mova.elamur.ai/free/en/docs/papers/mtrans/) / Verbmobil: Towards a DRT-based translation of spontaneous negotiation dialogues

# Verbmobil: Towards a DRT-based translation of spontaneous negotiation dialogues

**Authors:** Wolfgang Wahlster · **Year:** 1995 · **Venue:** Proceedings of Machine Translation Summit V **Link:** [aclanthology.org/1995.mtsummit-1.39](https://aclanthology.org/1995.mtsummit-1.39/) · **Licence:** [CC BY-NC-SA 3.0](https://creativecommons.org/licenses/by-nc-sa/3.0/) — summary only

[Proceedings of Machine Translation Summit V](https://aclanthology.org/volumes/1995.mtsummit-1/)[licence](https://creativecommons.org/licenses/by-nc-sa/3.0/)

- [Publisher's page](https://aclanthology.org/1995.mtsummit-1.39/)
- [PDF](https://aclanthology.org/1995.mtsummit-1.39.pdf)

**Authors (1)**

1. Wahlster, Wolfgang

Citations and author metrics from OpenAlex (); authors with the same name are not always told apart.

## What it is about

The paper describes the goals of the Verbmobil project, a speech translation effort on spontaneous negotiation dialogues, and the architecture of its first demonstrator in an appointment-scheduling domain. The translation component maps fully specified HPSG signs, with a source Discourse Representation Structure, onto an underspecified target DRS, and a generator based on a lexicalised tree-adjoining grammar produces the target text. The paper also discusses a prosodic module and a domain model in a description logic. The text file in the library holds only the abstract, so the body of the paper was not available to read. The summary above comes from the abstract alone.

## Ideas

- I1. Map a fully specified source structure onto an underspecified target one. In `tools/translate` the meta-world could leave out of a concept what the source fixes and the target does not need: Hebrew gender and aspect that English does not write. The encoder then fills those features from the target's own context counts. Test it: drop one source feature at a time from the meta tokens and measure whether the encoder's bits rise. A feature whose removal costs nothing belongs to the source alone.
- I2. A demonstrator in one narrow domain (appointment scheduling) before broad text. For Mova: report the translator's residual on narrow, repetitive chapters (the genealogies of 1 Chronicles 1–9, a training book; never the test books) next to the narrative chapters. That separates lexicon gaps from construction gaps.

## For Mova

Only the abstract was available (the body is not in the library), so the ideas come from its architecture sentence alone. I1 touches the meta tokens' features (open problem 5, number and person from FEATS, read in the other direction: which features the meta-world may drop).

Topics: [Machine translation](https://md.mova.elamur.ai/free/en/docs/papers/topics/machine-translation/)

## Cite

The paper itself, not this card. BibTeX from the library's record of the work.

```
@inproceedings{wahlster1995verbmobil,
  title = {{Verbmobil: Towards a DRT-based translation of spontaneous negotiation dialogues}},
  author = {Wahlster, Wolfgang},
  booktitle = {Proceedings of Machine Translation Summit V},
  year = {1995},
  url = {https://aclanthology.org/1995.mtsummit-1.39/}
}
```

- [← previous Developing an interlingual translation lexicon using WordNets and Grammatical Framework](https://md.mova.elamur.ai/free/en/docs/papers/virk-2014-developing-interlingual-translation-lexicon-wordnets/)
- [next → CharacTer: Translation Edit Rate on Character Level](https://md.mova.elamur.ai/free/en/docs/papers/wang-2016-character-translation-edit-rate-character/)
- [↑ back to Machine translation](https://md.mova.elamur.ai/free/en/docs/papers/mtrans/)