Designing an Automatic Sentence based Parsing for Ge’ez Sentences

This study addresses the need for parsing tools in Ge’ez natural language processing (NLP) systems. Parsing breaks down sentences into their component parts (words and phrases) and assigns grammatical roles. This research provides the basis for more complex NLP tasks by designing a parser specifically for declarative Ge’ez sentences. The parser utilizes context-free grammar (CFG) rules, a lexicon, and the Earley parsing algorithm, known for its effectiveness with context-free languages. The researchers evaluated the parser’s accuracy on a test set of 355 sentences, achieving an impressive 96.6% accuracy after applying error correction techniques. This work lays the way for further development of Ge’ez text parsers. Future advancements could involve expanding the parser’s capabilities to handle more intricate sentence structures and enrich its understanding of the language. By refining this tool, researchers can unlock the potential of Ge’ez for deeper analysis and broader NLP applications.