Lecture
This lecture covers coreference resolution, a natural language processing task that involves identifying when two or more expressions in text refer to the same entity. The instructor discusses the challenges of mention detection, different coreference resolution models, and the current state-of-the-art neural systems. The lecture also explores the applications of coreference resolution in full text understanding, machine translation, and dialogue systems. Various linguistic concepts related to coreference, anaphora, and cataphora are explained, along with traditional algorithms like Hobbs' naive algorithm. The lecture concludes with an overview of end-to-end neural coreference models and the latest advancements using BERT-based approaches.