Akkadian Treebank for early Neo-Assyrian Royal Inscriptions

What it is about

The paper describes the first Universal Dependencies treebank for Akkadian: 1,845 sentences and 22,277 words from the early Neo-Assyrian royal inscriptions (Grayson 1991), taken from the lemmatised Oracc edition. The corpus has no punctuation, so sentence boundaries were found from the annotation itself: the sentences are the tokens that have dependents and no head. Morphology was checked against the BabyFST analyser, which agreed with the manual lemmas for 94.6 percent of tokens and with the full analysis for 85.5 percent. The authors then tested the TurkuNLP parser on an 80/10/10 split. With segmentation cues it reached LAS 93.29 and MLAS 87.53; on unsegmented running text LAS fell to 69.95 and MLAS to 58.97. The paper also records its UD choices for Akkadian: no AUX, no PUNCT, construct state as nmod:poss, the relative particle ša as a relative clause, and the quotative mā as discourse.

Ideas

• I1. Sentence roots without punctuation: a token that has dependents and no head marks the start of a sentence in unpunctuated text. Use this as a boundary rule before parsing in world and prag, and test it on the unpunctuated Hebrew and Aramaic texts.
• I2. A two-condition measurement: run the Hebrew parser pipeline with and without sentence cues on the same text and report the gap, as the paper does (93 against 70 LAS). A large gap says the sentence splitter, not the parser, is the bottleneck.
• I3. Construct and relative-clause rules: the construct state as a possessive dependency of the head noun, and a relative clause as a dependent that follows the noun it qualifies. These are small rules over the trees, testable on a few hundred Hebrew noun phrases.

For Mova

I1 belongs in world and prag (sentence boundaries for unpunctuated input), and in tools/hbo, since the MACULA texts are also unpunctuated. I2 is a measurement for tools/hbo. I3 goes into The Noun, the Adjective and the Numeral and Nouns, adjectives, pronouns and numerals, and the label decisions are a reference for treebanks/akk.

Topics: Akkadian and cuneiform NLP · Treebanks

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/2020.tlt-1.11

@inproceedings{luukko2020akkadian,
  title = {{Akkadian Treebank for early Neo-Assyrian Royal Inscriptions}},
  author = {Luukko, Mikko and Sahala, Aleksi and Hardwick, Sam and Lindén, Krister},
  booktitle = {Proceedings of the 19th International Workshop on Treebanks and Linguistic Theories},
  year = {2020},
  doi = {10.18653/v1/2020.tlt-1.11},
  url = {https://doi.org/10.18653/v1/2020.tlt-1.11}
}

References

25 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Øyvind Bjøru and Na’ama Pat-El. 2020. The Historical Syntax of the Subordinative Morphemes in Assyrian Akkadian. Zeitschrift für Assyriologie und Vorderasiatische Archäologie, 110(1):71–83. find
2. Eran Cohen. 2012. Conditional Structures in Mesopotamian Old Babylonian (Languages of the Ancient Near East, 4). Eisenbrauns, Winona Lake, IN. find
3. Karlheinz Deller. 1957a. Zur sprachlichen Einordnung der Inschriften Aššurnaṣirpals II. (883–859). Orientalia Nova Series 26(2):144–156. find
4. Karlheinz Deller. 1957b. Assyrisches Sprachgut bei Tukulti-Ninurta II (888–884). Orientalia Nova Series 26(3):268–272. find
5. Guy Deutscher. 2000. Syntactic Change in Akkadian: The Evolution of Sentential Complementation. Oxford University Press, Oxford and New York. find
6. Dietz Otto Edzard. 2003. Sumerian Grammar (Handbuch der Orientalistik: Der Nahe und der Mittlere Osten, 71). Brill, Leiden and Boston. find
7. Eckart Frahm. 2019. The Neo-Assyrian Royal Inscriptions as Text: History, Ideology, and Intertextuality. In Writing Neo-Assyrian History: Sources, Problems, and Approaches (State Archives of Assyria Studies, 29), edited by Giovanni B. Lanfranchi, Raija Mattila, and Robert Rollinger, 139–159. The Neo-Assyrian Text Corpus Project, Helsinki. find
8. A. Kirk Grayson. 1991. Assyrian Rulers of the Early First Millennium B.C. I (1114 – 859 B.C.) (Royal Inscriptions of Mesopotamia, Assyrian Periods, 2). University of Toronto Press, Toronto. find
9. Jaakko Hämeen-Anttila. 2000. A Sketch of Neo-Assyrian Grammar (State Archives of Assyria Studies, 13). The Neo-Assyrian Text Corpus Project, Helsinki. find
10. John Huehnergard. 2 2005. A Grammar of Akkadian (Harvard Semitic Museum Studies, 45). Eisenbrauns, Winona Lake, IN. find
11. John Huehnergard and Chris Woods. 2008. Akkadian and Eblaite, In The Ancient Languages of Mesopotamia, Egypt and Aksum, edited by Roger D. Woodard. 83–153. Cambridge University Press, Cambridge. find
12. Mans Hulden. 2009. Foma: A Finite-State Compiler and Library. Proceedings of the 12th Conference of the European Chapter of the Association for Computational Linguistics (EACL): Demonstrations Session, 29–32. Association for Computational Linguistics, Athens. find
13. Guglielmo Inglese. 2015. Towards a Hittite Treebank. Basic Challenges and Methodological Remarks. In Proceedings of the Workshop on Corpus-Based Research in the Humanities (CRH) 10 December 2015, Warsaw, Poland, edited by Francesco Mambrini, Marco Passarotti, and Caroline Sporleder, 59–68. find
14. Jenna Kanerva, Filip Ginter, Niko Miekka, Akseli Leino, and Tapio Salakoski. 2018. Turku Neural Parser Pipeline: An End-to-End System for the CoNLL 2018 Shared Task. In Proceedings of the CoNLL 2018 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies, edited by Daniel Zeman and Jan Hajič, 133– 142. Association for Computational Linguistics, Brussels. find our card
15. N.J.C. Kouwenberg. 2017. A Grammar of Old Assyrian (Handbuch der Orientalistik, 1/118). Brill, Leiden and Boston. find
16. Krister Lindén, Miikka Silfverberg, and Tommi Pirinen. 2009. HFST tool for morphology: An efficient open-source package for construction of morphological analyzers. In State of the Art in Computational Morphology, edited by Cerstin Mahlow and Michael Piotrowski, 28–47. Communications in Computer and Information Science, Berlin and Heidelberg. find
17. Mikko Luukko. 2004. Grammatical Variation in Neo-Assyrian (State Archives of Assyria Studies, 16). The Neo-Assyrian Text Corpus Project, Helsinki. find
18. Erica Reiner. 1966. A Linguistic Analysis of Akkadian (Janua Linguarum, Series Practica, 21). Mouton, The Hague. find
19. Aleksi Sahala, Miikka Silfverberg, Antti Arppe, and Krister Lindén. 2020. BabyFST – Towards a Finite-State Based Computational Model of Ancient Babylonian. In Proceedings of the 12th Language Resources and Evaluation Conference (LREC), 3886–3894. find our card
20. SIGNLL = Special Interest Group on Natural Language Learning of the Association for Computational Linguistics. 2018. Evaluation script for the 2018 CoNLL shared task, web description. http://universaldependen-cies.org/conll18/evaluation.html ↗
21. Wolfram von Soden. 1973. Der akkadische Subordinativ-Subjunktiv. Zeitschrift für Assyriologie und Vorderasiatische Archäologie 63(1):56–58. find
22. Wolfram von Soden. 3 1995. Grundriss der akkadischen Grammatik (Analecta Orientalia, 33/47). Pontificium Institutum Biblicum, Rome. find
23. Michael P. Streck. 2010. Großes Fach Altorientalistik: Der Umfang des keilschriftlichen Textkorpus. Mitteilungen der Deutschen Orient-Gesellschaft, 142:35–58. find
24. Michael P. Streck. 2011. Babylonian and Assyrian. In The Semitic Languages: An International Handbook, edited by Stefan Weninger, 359–396. De Gruyter Mouton, Berlin and Boston. find
25. Imed Zitouni (ed.). 2014. Natural Language Processing of Semitic Languages. Springer, Berlin and Heidelberg. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/treebanks-2020-luukko-akkadian-treebank-early-neo-assyrian/