Tagging a Hebrew Corpus: the Case of Participles
What it is about
The authors build a part-of-speech and morphology corpus of Modern Hebrew news (about 40M tokens, with a 200K-token sample tagged by hand). Their guidelines start from dictionaries and the Penn Treebank scheme, and they revise them as taggers disagree. The main problem is the beinoni (participle) form, which shares properties of nouns, adjectives and verbs. Eight Hebrew sources assign it almost systematically different verb, noun or adjective tags. The authors add a separate participle tag, defined by morphological and syntactic tests: a present-verb use is tense-bound and needs an explicit subject, while noun and adjective uses take no tense and can carry a possessive suffix or a modifier. Disagreement between taggers fell from about 20% to 3% after a few rounds of revision. The tag also helped a syntactic chunker. The part of the paper read gives agreement figures and the tests, not a single accuracy headline.
Ideas
- I1. Keep participle as its own category, not forced into verb, noun or adjective. In the Modern Hebrew UD analysis, a beinoni form gets a participle feature, and the tree decides whether it works as a verb relation or as a nominal.
- I2. Turn the paper's diagnostics into rules: a present reading must survive substitution of past or future; a verbal use needs an explicit subject; a noun use can take a definite or possessive marker, and a verbal use cannot take the genitive marker šel with a noun-like structure. Each test is a few lines in the frame builder.
- I3. Build a small test set from the words in their Table 1 (the beinoni forms and the POS each of the eight sources gives). A lookup that reports conflicting POS across sources must flag exactly these words.
For Mova
I1 goes into 11. Modern Hebrew in Universal Dependencies and 05. The verb: system, binyanim, tense and use. I2 is a rule set for 05b. Verb paradigms: the strong verb in seven binyanim and the weak roots and the frame reader. I3 is a test set for the Modern Hebrew lexicon, 8. Lexicon and register.
Cite
@inproceedings{adler2008tagging,
title = {{Tagging a Hebrew Corpus: the Case of Participles}},
author = {Adler, Meni and Netzer, Yael and Goldberg, Yoav and Gabay, David and Elhadad, Michael},
booktitle = {Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08)},
year = {2008},
doi = {10.63317/5nvswopp4v97},
url = {https://doi.org/10.63317/5nvswopp4v97}
}References
- Meni Adler and Michael Elhadad. 2006. An unsupervised morpheme-based HMM for Hebrew morphological disambiguation. In Proceeding of COLING-ACL-06, Sydney, Australia. find our card
- Meni Adler. 2007. Hebrew Morphological Disambiguation: An Unsupervised Stochastic Word-based Approach. Ph.D. thesis, Ben-Gurion University of the Negev, Beer-Sheva, Israel. find
- Eitan Avneyon, Raphael Nir, and Idit Yosef. 2002. Milon sapir: The Encyclopedic Sapphire Dictionary. Hed Artsi, Tel-Aviv, Israel. (in Hebrew). find
- Yehoshua Blao. 1966. Syntax Fundamentals. Hebrew Institute for Written Education, Jerusalem. in Hebrew. find
- David Carmel and Yoelle S. Maarek. 1999. Morphological disambiguation for Hebrew search systems. In Proceeding of NGITS-99, pages 312–326. find
- Yaacov Choueka, Uzi Freidkin, Hayim A. Hakohen, , and Yael Zachi-Yannay. 1997. Rav Milim: A Comprehensive Dictionary of Modern Hebrew. Steimatsky, Tel-Aviv, Israel. (in Hebrew). find
- Herve Dejean. 2000. How to evaluate and compare tagsets? a proposal. In Proceedings of LREC 2000, Athens, Greece. find
- Edit Doron. 2000. The passive participle. Hebrew Linguistics, 47:39–62. (in Hebrew). find
- Avraham Even-Shoshan. 2003. Even Shoshan’s Dictionary - Renewed and Updated for the 2000s. Am Oved, Kineret, Zmora-Bitan, Dvir and Yediot Aharonot. (in Hebrew). find
- Friedrich H. W. Gesenius. 1976. Hebrew Grammar. The Clarendon Press, Oxford. Edited and enlarged by E. Kautzsch, English edition by A. E. Cowley. find
- Yoav Goldberg, Michael Elhadad, and Meni Adler. 2006. Noun phrase chunking in hebrew influence of lexical and morphological features. In Proceeding of COLING-ACL-06, Sydney, Australia. find our card
- Yaakov Knaani. 1960. The Hebrew Language Lexicon. Masada, Jerusalem, Israel. (in Hebrew). find
- Taku Kudo and Yuji Matsumato. 2000. Use of support vector learning for chunk identification. In Proceedings of CoNLL-00 and LLL-00, Lisbon, Portugal. find
- Mitchell P. Marcus, Beatrice Santorini, and Mary Ann Marchinkiewicz. 1993. Building a large annotated corpus of English: The penn treebank. Computational Linguistics, 19:313–330. find our card
- Mark Van Mol. 2002. The semi-automatic tagging of arabic corpora. In Arabic Language Resources and Evaluation - Status and Prospects Workshop, LREC. find
- Yael Netzer, Meni Adler, David Gabay, and Michael ELhadad. 2007. Can you tag the modal? you should! In ACL07 Workshop on Computational Approaches to Semitic Languages, Prague, Czech. find
- Uzi Ornan. 2002. Hebrew in Latin script. Lĕšonénu, LXIV:137–151. (in Hebrew). find
- Haim B. Rosen. 1977. Contemporary Hebrew. Mouton, The Hague, Paris. find
- Beatrice Santorini. 1995. Part-of-speech tagging guidelines for the Penn Treebank Project. 3rd revision;. Technical report, Department of Computer and Information Science, University of Pennsylvania. find our card
- Erel Segal. 2000. Hebrew morphological analyzer for Hebrew undotted texts. Master’s thesis, Technion, Haifa, Israel. (in Hebrew). find
- Ur Shlonsky. 1997. Clause Structure and Word Order in Hebrew and Arabic. Oxford Universirt Press, New York Oxford. find
- Khalil Sima’an, Alon Itai, Alon Altman Yoad Winter, and Noa Nativ. 2001. Building a tree-bank of modern Hebrew text. Journal Traitement Automatique des Langues (t.a.l.). Special Issue on NLP and Corpus Linguistics. find
- Shlomo Yona. 2004. A finite-state based morphological analyzer for Hebrew. Master’s thesis, Haifa University. find
This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/lingbulk-2008-adler-tagging-hebrew-corpus-case-participles/