We are pleased to announce that MA student Anna Stacey has successfully defended her thesis titled “Automatic Glossing for Northern Interior Languages Featuring Cross-Lingual Enhancements.” Her research was supervised by Dr. Miikka Silfverberg, with committee members Dr. Henry Davis and Dr. Garrett Nicolai.
Anna’s project addresses the urgent need for efficient language documentation by exploring how cross-lingual transfer techniques can improve automatic glossing for low-resource languages. Focusing on three Northern Interior Salish languages—St’át’imcets, nɬeʔkepmxcín, and Secwepemctsín—her work leverages machine learning methods to improve glossing performance, particularly where data is scarce.
By combining multilingual training with enhancements such as fine-tuning and data augmentation, Anna demonstrated improved performance for languages with limited resources. Notably, her research shows the potential of zero-shot learning in segmenting previously unseen words, highlighting the promise of cross-lingual approaches for supporting Indigenous language revitalization.
Congratulations, Anna, on this important and impactful contribution to computational linguistics and language documentation!