Parsing Akkadian Verbs with Prolog

What it is about

A Prolog system that parses and generates finite Akkadian verb forms, with optional accusative, dative and ventive suffixes, for the G, D and N stems. The grammar is written as definite clause grammar rules (about 200 per stem), and the same rules both parse and generate. Vowels are mostly ignored; where a vowel's length is unknown it can be given with an asterisk, and the system expands it to the three lengths, so one input returns several readings (the paper's example gives a G durative, an N preterite and an N durative). Each stem is tried with and without a suffix. In April 2002 the strong and singly weak triradical G, D and N stems were covered, behind a web interface that handled about ten forms per second. The author states that a two-stage rule approach was too intricate for weak verbs. No accuracy figures against a corpus are given.

Ideas

• I1. A vowel wildcard for unpointed Hebrew input: where a vowel or its length is unknown, the segmenter enumerates the few possible vowels for that slot, not all strings. Add it to the unpointed mode of tools/hbo and report the number of pointings per token.
• I2. Use the grammar both ways as a round-trip test: generate every form of a lexeme from the paradigm, parse each one back, and check that the lexeme and the features return. Run it on the Hebrew verb paradigms in treebanks/hbo/ptnk.
• I3. Parse the stem first, then run the object-suffix grammar on the remainder, with and without the suffix (the paper's two-pass method). For Hebrew verbs with pronominal objects this is the same two-pass design.

For Mova

I1 goes into the unpointed mode of tools/hbo. I2 is a round-trip test for The verb. I3 goes into Pronouns, particles and prepositions. The Akkadian grammar itself is not in Mova; the ideas are the method.

Topics: Akkadian and cuneiform NLP

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.3115/1118637.1118638

@inproceedings{macks2002parsing,
  title = {{Parsing Akkadian Verbs with Prolog}},
  author = {Macks, Aaron},
  booktitle = {Proceedings of the ACL-02 Workshop on Computational Approaches to Semitic Languages},
  year = {2002},
  doi = {10.3115/1118637.1118638},
  url = {https://doi.org/10.3115/1118637.1118638}
}

References

5 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Buccellati, Giorgio. A Structural Grammar of Babylonian. Weisbaden: Harrassowitz Verlag. 1996. Hall, Marty Guide to CGI Programming in Java. http://www.apl.jhu.edu/~hall/java/CGI-with-Java.html. 1998 Huehnergard, John. A Grammar of Akkadian. ↗
2. Atlanta, Georgia: Scholars Press. 1997. find
3. Kataja , Laura and K. Koskenniemi. 1988). find
4. “Finite-state description of Semitic morphology: a case study in Ancient Akkadian.” Proceedings of the International Conference on Computational Linguistics (COLING88), pages 313--315. Miller, Douglas B An Akkadian Handbook. Winona Lake, Ind. Eisenbrauns, 1996. Richardson, M. E. J. H a m m u r a b i ’ s Laws:Text, Translation and Glossary. Sheffield, England: Sheffield Academic Press. 2000. Von Soden, Wolfram and Werner R. Meyer. find
5. Grundriss der Akkadischen Grammatik. Rome : Editrice Pontificio Istituto Biblico. 1995. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/aclbulk-2002-macks-parsing-akkadian-verbs-prolog/