Explore document-level models in natural language processing through this comprehensive lecture on coreference resolution and discourse parsing. Dive into various coreference models, including mention pair and entity models, and understand the advantages of neural network approaches. Learn about document-level language modeling, methods for evaluating document coherence, and techniques for mention detection. Examine end-to-end neural coreference systems and their applications in neural models. Investigate discourse parsing, including shift-reduce parsing and attention-based hierarchical neural networks. Discover how to leverage discourse structure in neural models to enhance NLP tasks.
Overview
Syllabus
Intro
Some Connections to Tasks over Documents
Document Level Language Modeling
What Context to Incorporate?
How to Evaluate Document Coherence Models?
Mention(Noun Phrase) Detection
Components of a Coreference Model . Like a traditional machine learning model
Coreference Models:Instances
Mention Pair Models
Entity Models
Advantages of Neural Network Models for Coreference
Coreference Resolution w/ Entity- Level Distributed Representations
End-to-End Neural Coreference (Span Model)
End-to-End Neural Coreference (Coreference Model)
Using Coreference in Neural Models
Document Problems: Discourse Parsing
Shift-reduce Parsing Discourse Structure Parsing w/ Distributed Representations (Ji and Eisenstein 2014) . Shift-reduce parser with features from 2 stack elements and queue element
Discourse Parsing w/ Attention- based Hierarchical Neural Networks
Uses of Discourse Structure in Neural Models
Taught by
Graham Neubig
