Overview
Learn about cross-lingual approaches to named entity recognition in this 53-minute lecture presented by Junjie Hu at USC Information Sciences Institute. Explore how to extend NLP systems' ability to understand and generate named entities across different languages. Examine the challenges of annotating data for all real-world scenarios and the limitations of current deep learning models when dealing with multilingual content. Discover methods for improving named entity handling in cross-lingual applications like task-oriented dialog and machine translation. Gain insights from Hu's research at the intersection of NLP and machine learning, including work on multilingual NLP, transfer learning, and knowledge representation learning.
Syllabus
Cross-lingual Learning of Named Entities
Taught by
USC Information Sciences Institute