MRL Parsing Without Tears: The Case of Hebrew

What it is about

Parsers for morphologically rich languages either run a pipeline (segment, tag, parse), which passes errors forward, or a joint lattice over all segmentations, which is slow and needs a lexicon. The authors propose a "flipped pipeline" on DictaBERT: each whole space-delimited token gets expert classifiers on shared BERT embeddings — dependency head and label (MST at inference), lemma, morphology (POS, features, proclitic functions, suffix and its features), proclitic segmentation (8 letter-group classes, limited to valid sets) and NER. Only at the end is a UD tree synthesised: prefix and suffix tokens are split out and given lemmas and relations by rules. No lexicon is used. Trained on HTB and NEMO it is near the state of the art; with IAHLT data added it reaches a new state of the art for Modern Hebrew POS tagging and dependency parsing, released as DictaBERT-Parse in three sizes.

Ideas

• I1. Test the error-propagation claim on tools/hbo: count how many dependency errors come from a wrong segmentation, and compare with a whole-token head decision made before segmentation.
• I2. Synthesis rules for the CoNLL-U output, as in the paper's Figure 1 and appendix: the prefix lemma is its letters, the prefix relation follows from the host token, suffix pronouns are split out last. Write them as a deterministic table and test them on MACULA.
• I3. Lexicon-free check: measure the segmenter on words outside its lex table and on foreign or code-switched words, with and without the table, and report the drop.

For Mova

I2 goes into tools/hbo (the CoNLL-U writer after gparse) and into Syntax: the clause and the sentence for prefix attachment; for Modern Hebrew it is the design in 11. Modern Hebrew in Universal Dependencies and 12-mova.md. I1 and I3 are measurements for tools/hbo.

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/2024.findings-acl.269

@inproceedings{shmidman2024mrl,
  title = {{MRL Parsing Without Tears: The Case of Hebrew}},
  author = {Shmidman, Shaltiel and Shmidman, Avi and Koppel, Moshe and Tsarfaty, Reut},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2024},
  year = {2024},
  doi = {10.18653/v1/2024.findings-acl.269},
  url = {https://doi.org/10.18653/v1/2024.findings-acl.269}
}

References

22 references, 16 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Dan Bareket and Reut Tsarfaty. 2021. Neural modeling for named entities and morphology (NEMO2). Transactions of the Association for Computational Linguistics, 9:909–928. find our card
2. Shay B. Cohen and Noah A. Smith. 2007. Joint morphological and syntactic disambiguation. In Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (EMNLP-CoNLL), pages 208–217, Prague, Czech Republic. Association for Computational Linguistics. find
3. Marie-Catherine de Marneffe, Christopher D. Manning, Joakim Nivre, and Daniel Zeman. 2021. Universal Dependencies. Computational Linguistics, 47(2):255–308. find our card
4. Matan Eyal, Hila Noga, Roee Aharoni, Idan Szpektor, and Reut Tsarfaty. 2023. Multilingual sequence-to-sequence models for Hebrew NLP. In Findings of the Association for Computational Linguistics: ACL 2023, pages 7700–7708, Toronto, Canada. Association for Computational Linguistics. find our card
5. Yoav Goldberg and Reut Tsarfaty. 2008. A single generative model for joint morphological segmentation and syntactic parsing. In Proceedings of ACL-08: HLT, pages 371–379, Columbus, Ohio. Association for Computational Linguistics. find
6. Omer Goldman and Reut Tsarfaty. 2022. Morphology without borders: Clause-level morphology. Transactions of the Association for Computational Linguistics, 10:1455–1472. find our card
7. Eliyahu Kiperwasser and Yoav Goldberg. 2016. Simple and accurate dependency parsing using bidirectional LSTM feature representations. Transactions of the Association for Computational Linguistics, 4:313– 327. find our card
8. Amit Seker, Elron Bandel, Dan Bareket, Idan Brusilovsky, Refael Shaked Greenfeld, and Reut Tsarfaty. 2021. Alephbert:a hebrew large pre-trained language model to start-off your hebrew nlp application with. find our card
9. Stav Klein and Reut Tsarfaty. 2020. Getting the find
10. Amit Seker and Reut Tsarfaty. 2020. A pointer network architecture for joint morphological segmentation and tagging. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 4368–4378, Online. Association for Computational Linguistics. find our card
11. Amrith Krishna, Ashim Gupta, Deepak Garasangi, Pavankumar Satuluri, and Pawan Goyal. 2020a. Keep it surprisingly simple: A simple first order graph based parsing model for joint morphosyntactic parsing in Sanskrit. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 4791–4797, Online. Association for Computational Linguistics. find
12. Shaltiel Shmidman, Avi Shmidman, and Moshe Koppel. 2023. Dictabert: A state-of-the-art bert suite for modern hebrew. find our card
13. Reut Tsarfaty. 2006. Integrated morphological and syntactic disambiguation for Modern Hebrew. In Proceedings of the COLING/ACL 2006 Student Research Workshop, pages 49–54, Sydney, Australia. Association for Computational Linguistics. find our card
14. Danit Yshaayahu Levi and Reut Tsarfaty. 2024. A truly joint neural architecture for segmentation and parsing. find our card
15. Reut Tsarfaty, Shoval Sadde, Stav Klein, and Amit Seker. 2019. What’s wrong with Hebrew NLP? and how to make it right. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP): System Demonstrations, pages 259–264, Hong Kong, China. Association for Computational Linguistics. find our card
16. Amir More, Amit Seker, Victoria Basmova, and Reut Tsarfaty. 2019. Joint transition-based models for morpho-syntactic parsing: Parsing strategies for MRLs and a case study from Modern Hebrew. Transactions of the Association for Computational Linguistics, 7:33–48. find our card
17. Shaked Yehezkel and Yuval Pinter. 2023. Incorporating context into subword vocabularies. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 623–635, Dubrovnik, Croatia. Association for Computational Linguistics. find
18. Amrith Krishna, Bishal Santra, Ashim Gupta, Pavankumar Satuluri, and Pawan Goyal. 2020b. A graph-based framework for structured prediction tasks in Sanskrit. Computational Linguistics, 46(4):785–845. find
19. Minh Van Nguyen, Viet Dac Lai, Amir Pouran Ben Veyseh, and Thien Huu Nguyen. 2021. Trankit: A light-weight transformer-based toolkit for multilingual natural language processing. find our card
20. Peng Qi, Yuhao Zhang, Yuhui Zhang, Jason Bolton, and Christopher D. Manning. 2020. Stanza: A python natural language processing toolkit for many human languages. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 101–108, Online. Association for Computational Linguistics. find our card
21. Shoval Sade, Amit Seker, and Reut Tsarfaty. 2018. The Hebrew Universal Dependency treebank: Past present and future. In Proceedings of the Second Workshop on Universal Dependencies (UDW 2018), pages 133–143, Brussels, Belgium. Association for Computational Linguistics. find our card
22. Wolfgang Seeker and Özlem Çetinoğlu. 2015. A graph-based lattice dependency parser for joint morphological segmentation and syntactic analysis. Transactions of the Association for Computational Linguistics, 3:359–373. find our card

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/findings-2024-shmidman-mrl-parsing-without-tears-case/