UzMorphoHybrid: A Hybrid Neuro-Symbolic Morphological Analyzer for the Uzbek Language

Key Points

• This research aims to develop UzMorphoHybrid, a novel morphological analyzer specifically designed for the Uzbek language.
• Developed an open-source hybrid morphological analyzer for Uzbek language processing.
• Integrated neuro-symbolic approaches to improve analysis of low-frequency word forms.
• Compared performance with existing statistical models.
• UzMorphoHybrid shows improved accuracy in analyzing rare word forms compared to traditional models.
• The hybrid approach effectively reduces errors in morphological parsing, particularly for agglutinative structures.
• User feedback indicates greater satisfaction with UzMorphoHybrid's performance.