🤖 AI Summary
Ge’ez, a highly inflectional Classical Semitic language, suffers from severe scarcity of annotated morphological data. To address this low-resource challenge, this work introduces the first rule-based morphological synthesizer for Ge’ez augmented with Transformation Learning (TLM). The system covers all verbal inflectional paradigms using a lexicon of 1,102 verbs and accurately generates surface forms from lexical roots. Unlike data-hungry statistical approaches, it combines interpretable linguistic rules with TLM-driven optimization to maximize generalization under minimal supervision. Evaluated on a standard test set, the synthesizer achieves 97.4% accuracy—substantially outperforming baseline models. This work establishes the first publicly available morphological synthesizer for Ge’ez, filling a critical gap in Ge’ez NLP tooling. Moreover, its hybrid rule–learning methodology provides a reusable, extensible framework for morphological analysis of other low-resource classical languages.
📝 Abstract
Ge’ez is an ancient Semitic language renowned for its unique alphabet. It serves as the script for numerous lan- guages, including Tigrinya and Amharic, and played a pivotal role in Ethiopia’s cultural and religious development during the Aksumite kingdom era. Ge’ez remains significant as a liturgical language in Ethiopia and Eritrea, with much of the national identity documentation recorded in Ge’ez. These written materials are invaluable primary sources for studying Ethiopian and Eritrean philosophy, creativity, knowledge, and civilization. Ge’ez is a complex morphological structure with rich inflectional and derivational morphology, and no usable NLP has been developed and published until now due to the scarcity of annotated linguistic data, corpora, labeled datasets, and lexicons. Therefore, we proposed a rule-based Ge’ez morphological synthesis to generate surface words from root words according to the morphological structures of the language. Consequently, we proposed an automatic morphological synthesizer for Ge’ez using TLM. We used 1,102 sample verbs, representing all verb morphological structures, to test and evaluate the system. Finally, we get a performance of 97.4%. This result outperforms the baseline model, suggesting that other scholars build a comprehensive system considering morphological variations of the language. Keywords: Ge’ez, NLP, morphology, morphological synthesizer, rule-based