--- title: "LIMSI@CoNLL'17: UD Shared Task — Aufrant 2017" description: "The paper describes the LIMSI team's system for the CoNLL 2017 UD Shared Task." authors: ["Lauriane Aufrant", "Guillaume Wisniewski", "François Yvon"] year: 2017 venue: "Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies" source: "https://aclanthology.org/K17-3017/" doi: "10.18653/v1/k17-3017" licence: "CC BY 4.0 (ACL Anthology)" licence_url: "https://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/papers/aufrant-2017-limsi-conll-17-ud-shared/" --- [Elamur Research](https://md.research.elamur.ai/) · [this page in HTML](https://research.elamur.ai/free/papers/aufrant-2017-limsi-conll-17-ud-shared/) [home](https://md.research.elamur.ai/) / [papers](https://md.research.elamur.ai/free/papers/) / [2017](https://md.research.elamur.ai/free/papers/2017/page/2/) / LIMSI@CoNLL'17: UD Shared Task # LIMSI@CoNLL'17: UD Shared Task **Authors:** [Lauriane Aufrant](https://md.research.elamur.ai/free/authors/lauriane-aufrant/), [Guillaume Wisniewski](https://md.research.elamur.ai/free/authors/guillaume-wisniewski/), [François Yvon](https://md.research.elamur.ai/free/authors/francois-yvon/) · **Year:** 2017 · **Venue:** Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/K17-3017/) · [PDF](https://aclanthology.org/K17-3017.pdf) · [DOI 10.18653/v1/k17-3017](https://doi.org/10.18653/v1/k17-3017) · [OpenAlex](https://openalex.org/W2740964521) **Citations:** 6 citations (openalex-snapshot · 2026-10-14) · [4](#cited-by) in our library, of the 140673 papers whose bibliographies we read ## Summary The paper describes the LIMSI team's system for the CoNLL 2017 UD Shared Task. This is a multilingual dependency parsing competition that goes from raw text to Universal Dependencies annotation. The authors focused on small treebanks with fewer than 1,000 training sentences. For these, they proposed a cascaded combination of several base parsers: monolingual, delexicalized cross-lingual, and trained on trees transferred through parallel corpora. Each parser is assigned a "competence zone" in advance, i.e., the subset of tokens it annotates. The parser takes into account the partial tree produced by the preceding stages, so each token is handled by exactly one parser. The best strategy for each language was selected on the dev data. On several small treebanks this yielded a noticeable gain over UDPipe. However, the overall result fell below the baseline because of sharp failures on certain languages. Analysis of these failures revealed substantial annotation discrepancies between treebanks of the same language. This is an important caveat for anyone combining several UD corpora of the same language. ## Key points - The system's official result is 67.72 LAS, ranking 17th out of 33 participants. The UDPipe 1.1 baseline obtained 68.35 LAS and ranked 13th. Had the same models as the baseline been submitted for the problematic treebanks, the system would have scored 68.90 LAS and ranked 9th. In the unofficial post-evaluation ranking it is 12th. - The core of the system is the authors' own parser, PanParser. It is transition-based: it uses ArcEager, an averaged perceptron, dynamic oracles, and a beam of size 8 or greedy search. PanParser can learn from partial trees and build parses subject to constraints, and this is what the cascade relies on. The UDPipe models in the cascade are not retrained; their output is only filtered. - The source language for cross-lingual transfer is chosen among treebanks with more than 2,000 sentences. They are first transformed with rules based on WALS, and then the one with the lowest KL cpos3 divergence is selected. Parallel data are taken from OPUS, aligned with fast_align, and partial projection following Lacroix et al. (2016) is applied. - The authors attribute the failures on the PUD treebanks to the fact that different corpora of the same language apply the annotation scheme differently. Moreover, in some corpora the "raw" text has already been partially preprocessed. In Russian-SynTagRus, Finnish-FTB and Greek, multiword tokens are already marked. Danish and Finnish-FTB are provided already tokenized. - The authors acknowledge the method's limitations: there are no confidence estimation mechanisms, delexicalized models are unreliable, and there are too few dev data to determine the competence zones accurately. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* Fields: [Linguistics & languages](https://md.research.elamur.ai/free/shelves/linguistics/) Topics: [Universal Dependencies](https://md.research.elamur.ai/free/topics/universal-dependencies/) · [Treebanks](https://md.research.elamur.ai/free/topics/treebanks/) ## Cited by Cited by 4 papers of our library cite it (of the 140673 whose bibliographies we read); 2 of them have a page here, 2 not yet. - [CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies](https://md.research.elamur.ai/free/papers/ud-2017-zeman-conll-2017-shared-task-multilingual/) - [Errator: a Tool to Help Detect Annotation Errors in the Universal Dependencies Project](https://md.research.elamur.ai/free/papers/ud-2018-wisniewski-errator-tool-help-detect-annotation/) ## Cite The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{aufrant2017limsi, title = {{LIMSI@CoNLL'17: UD Shared Task}}, author = {Aufrant, Lauriane and Wisniewski, Guillaume and Yvon, François}, booktitle = {Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies}, year = {2017}, doi = {10.18653/v1/k17-3017}, url = {https://doi.org/10.18653/v1/k17-3017} } ``` - [← previous Lexically Constrained Decoding for Sequence Generation Using Grid Beam Search](https://md.research.elamur.ai/free/papers/hokamp-2017-lexically-constrained-decoding-sequence-generation/) - [next → Linguistic realisation as machine translation: Comparing different MT models for AMR-to-text generation](https://md.research.elamur.ai/free/papers/ferreira-2017-linguistic-realisation-as-machine-translation/) - [↑ back to papers of 2017](https://md.research.elamur.ai/free/papers/2017/page/2/) ## References 4 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Ethem Alpaydin and Cenk Kaynak. 1998. Cascading classifiers. Kybernetika pages 369–374. [find](https://scholar.google.com/scholar?q=Ethem+Alpaydin+and+Cenk+Kaynak%2E+1998%2E+Cascading+classifiers%2E+Kybernetika+pages+369%E2%80%93374%2E) 2. Lauriane Aufrant and Guillaume Wisniewski. 2016. PanParser: a Modular Implementation for Efficient Transition-Based Dependency Parsing. Technical report, LIMSI-CNRS. [find](https://scholar.google.com/scholar?q=Lauriane+Aufrant+and+Guillaume+Wisniewski%2E+2016%2E+PanParser%3A+a+Modular+Implementation+for+Efficient+Transition%2DBased+Dependency+Parsing%2E+Technical+report%2C+LIMSI%2DCNRS%2E) 3. Lauriane Aufrant, Guillaume Wisniewski, and François Yvon. 2016. Zero-resource Dependency Parsing: Boosting Delexicalized Cross-lingual Transfer with Linguistic Knowledge. In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers. [find](https://scholar.google.com/scholar?q=Lauriane+Aufrant%2C+Guillaume+Wisniewski%2C+and+Franc%CC%A7ois+Yvon%2E+2016%2E+Zero%2Dresource+Dependency+Parsing%3A+Boosting+Delexicalized+Cross%2Dlingual+Transfer+with+Linguistic+Knowledge%2E+In+Proceedings+of+COLING+20) 4. The COL-ING 2016 Organizing Committee, Osaka, Japan, pages 119–130. [http://aclweb.org/anthology/C16-1012](http://aclweb.org/anthology/C16-1012). \#\# Conclusion [↗](http://aclweb.org/anthology/C16-1012)