A Transformer-Based Parser for Syriac Morphology

A Transformer-Based Parser for Syriac MorphologySyriaca.org: The Syriac Reference PortalThe National Endowment for the HumanitiesThe International Balzan Prize FoundationDavid A. Michelsoncbsc_adminRecord added to Zotero bycbsc_adminSyriaca.org: The Syriac Reference Portalhttp://syriaca.org/cbss/M6L9MI3X/tei

Distributed under a Creative Commons Attribution 3.0 Unported License.

2024-08-15-04:00

Born digital.

CREATED: This bibl record was autogenerated from a Zotero record.MartijnNaaijerConstantijn J.SikkelMathiasCoeckelbergsJiskAttemaWido Thvan PeursenA Transformer-Based Parser for Syriac Morphologyhttp://syriaca.org/cbss/M6L9MI3Xhttps://www.zotero.org/groups/a_comprehensive_bibliography_on_syriac_studies/items/M6L9MI3Xhttps://www.zotero.org/groups/4861694/items/M6L9MI3XProceedings of the Ancient Language Processing Workshop Associated with the 14th International Conference on Recent Advances in Natural Language Processing, RANLP 2023, 8 September, 2023ShoumenIncoma Ltd.202323-29In this project we train a Transformer-based model from scratch, with the goal of parsing the morphology of Ancient Syriac texts as accurately as possible. Syriac is a low-resource language, only a relatively small training set was available. Therefore, the training set was expanded by adding Biblical Hebrew data to it. Five different experiments were done: the model was trained on Syriac data only, it was trained with mixed Syriac and (un)vocalized Hebrew data, and it was trained first on (un)vocalized Hebrew data and then trained further on Syriac data. The models trained on Hebrew and Syriac data consistently outperform the models trained on Syriac data only. This shows that the differences between Syriac and Hebrew are small enough that it is worth adding Hebrew data to train the model for parsing Syriac morphology. Training models with data from multiple languages is an important trend in NLP, we show that this works well for relatively small datasets of Syriac and Hebrew. Naaijer, Martijn, Constantijn J. Sikkel, Mathias Coeckelbergs, Jisk Attema, and Wido Th van Peursen. “A Transformer-Based Parser for Syriac Morphology.” In Proceedings of the Ancient Language Processing Workshop Associated with the 14th International Conference on Recent Advances in Natural Language Processing, RANLP 2023, 8 September, 2023, 23–29. Shoumen: Incoma Ltd., 2023. Martijn Naaijer et al., A Transformer-Based Parser for Syriac Morphology, in Proceedings of the Ancient Language Processing Workshop Associated with the 14th International Conference on Recent Advances in Natural Language Processing, RANLP 2023, 8 September, 2023 (Shoumen: Incoma Ltd., 2023), 23–29.Syriac studiesXIX-XXI CEXXI CEdigital