Schedule

Overall Structure

Each course meeting will include some if not all of the following components:

• Assigned Reading - These papers must be read before the course, some students will give short presentations on the reading.
• Suggested Reading - These are optional papers that are related to the topic of the class meeting and
• Lecture / Discussion - Given by the instructor, TAs, or guest speaker on the subject area of the course meeting.
• Scribes - Students assigned to take notes and share notes with all students in the course.
• Course project assignments / updates - Assignments to be handed in electronically such as project proposals, midterm reports.

Typical Course Meeting

• Three 15-minute paper presentations on the assigned paper reading. Done in pairs of students. (~45 minutes)
• Three 3-minute spotlight presentations on the presenter's choice of paper from the suggested reading. Done individually. (~9 minutes)
• Lecture & discussion done by instructors or PhD students working on relevant areas (~1hr)
• Scribe duties: For each 15-minute presentation and for the course lecture, two students will be asked to perform scribe duties. These students write notes individually and both sets of notes will be shared on the course website. For each 3-minute spotlight presentation, one student will be asked to perform scribe duties. These notes will also be shared on the course website.

Reading assignments:

https://docs.google.com/spreadsheets/d/1b3Tn5hsQmyl65yMXWCY01FNpYVcF3YIlnFEvauZG34Q/edit?usp=sharing

Scribe assignments:

https://docs.google.com/spreadsheets/d/1QE2yMCmrYv2YghyAdJbKm8a9Le-ektOJ6i-wmndvPpA/edit#gid=0

Meeting 1 (9.6.2019)

Introduction to KBs, automated methods for KB construction

• Welcome!
• No assigned reading
• Course overview slides:

Meeting 2 (9.13.2019)

Knowledge Representations; Course Project Overview [KR]

Assigned Readings (see this for more information about paper presentation assignments)

• [KR:P1] Cyc: toward programs with common sense. Douglas B. Lenat, Ramanathan V. Guha, Karen Pittman, Dexter Pratt, Mary Shepherd. Commun. ACM. 1990. Presented by Dongxu Zhang & Shib Shankar Dasgupta [Slides] [Scribe 1] [Scribe 2]
• [KR:P2] Learning distributed representations of concepts. Proceedings of the eighth annual conference of the cognitive science society. 1986. Presented by Dung Thai and Rico Angell [Slides] [Scribe 1] [Scribe 2]
• [KR:P3] Identity. Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/identity/ Presented by Ameya Godbole and Ziqiang Guan [Slides] [Scribe 1]
• [KR:Spotlight1] Methods for Domain-Independent Information Extraction from the Web: An Experimental Comparison. Oren Etzioni, Michael J. Cafarella, Doug Downey, Ana-Maria Popescu, Tal Shaked, Stephen Soderland, Daniel S. Weld, Alexander Yates. AAAI 2004. Presented by Ryan Minichiello [Slides] [Scribe 1]
• [KR:Spotlight2] Ship of Theseus. https://en.wikipedia.org/wiki/Ship_of_Theseus (See also in De Corpore, 2.11) Presented by Evan Rourke [Slides]
• [KR:Spotlight3] TextRunner: Open Information Extraction on the Web. Presented by Raghuveer Thirukovalluru [Slides] [Scribe 1] [Scribe 2]

Meeting 3 (9.20.2019)

Mention Segmentation, Entity Typing, Entity Linking [Type+Link]

Lecture [Slides] [Scribe 1] [Scribe 2]

Assigned Readings:

• [Type+Link:P1] Neural Architectures for Named Entity Recognition. Guillaume Lample, Miguel Ballesteros, Sandeep Subramanian, Kazuya Kawakami, Chris Dyer · HLT-NAACL · 2016 and Design Challenges and Misconceptions in Named Entity Recognition. Lev-Arie Ratinov, Dan Roth · CoNLL · 2009 Presented by Sreeparna Mukherjee & Evan Rourke [Slides] [Scribe 1]
• [Type+Link:P2] DeepType: Multilingual Entity Linking by Neural Type System Evolution. Jonathan Raiman, Olivier Raiman · AAAI · 2018 Presented by Aartika Rai & Shruti Jalan [Slides] [Scribe 1] [Scribe 2]
• [Type+Link:P3] Deep Joint Entity Disambiguation with Local Neural Attention. Octavian-Eugen Ganea, Thomas Hofmann · EMNLP · 2017 Presented by Weiqiu You & Kriti Myer [Slides] [Scribe 1]
• [Type+Link:Spotlight1] Unsupervised Recurrent Neural Network Grammars. Presented by Dhruvesh Patel
• [Type+Link:Spotlight2] Zero-Shot Entity Linking by Reading Entity Descriptions. Presented by Ameya Godbole

Meeting 4 (9.27.2019)

Entity Resolution & Clustering [Clustering]

Lecture [Slides]

Assigned Reading

• [Clustering:P1] End-to-end Neural Coreference Resolution. Kenton Lee, Luheng He, Mike Lewis, Luke Zettlemoyer. EMNLP. 2017. Presented by Subendhu Rongali & Raghuveer Thirukovalluru [Slides] [Scribe 1] [Scribe 2]
• [Clustering:P2] Robust Entity Clustering via Phylogenetic Inference. Nicholas Andrews, Jason Eisner, Mark Dredze. ACL 2014. Presented by Ahsaas Bajaj & Shruti Jalan [Slides] [Scribe 1] [Scribe 2]
• [Clustering:P3] Affinity Clustering: Hierarchical Clustering at Scale. Mohammadhossein Bateni, Soheil Behnezhad, Mahsa Derakhshan, MohammadTaghi Hajiaghayi, Raimondas Kiveris, Silvio Lattanzi, Vahab Mirrokni. NeurIPS 2017. Presented by Ryan Minichiello & Andrew Drozdov [Slides] [Scribe 1]
• [Clustering:Spotlight1] Hierarchical Clustering better than Average-Linkage. Presented by Clayton Thorrez
• [Clustering:Spotlight2] Entity Clustering Across Languages. Presented by Weiqiu You
• [Clustering:Spotlight3] Joint Entity and Event Coreference Resolution across Documents. Presented by Bhanu Pratap Singh

Meeting 5 (10.04.2019)

Relation Extraction, Semantic Role Labeling, & Frames [RE]

Lecture [Slides 1] [Slides 2] [Scribe 1] [Scribe 2]

Assigned Reading

1. [RE:P1] Document-Level N-ary Relation Extraction with Multiscale Representation Learning. Robin Jia, Cliff Wong, Hoifung Poon, ACL 2019 and Simultaneously Self-Attending to All Mentions for Full-Abstract Biological Relation Extraction. Patrick Verga, Emma Strubell, Andrew McCallum. NAACL-HLT. 2018.
2. [RE:P2] Training Classifiers with Natural Language Explanations. Braden Hancock, Paroma Varma, Stephanie Wang, Martin Bringmann, Christopher Re´. ACL. 2018 and Joint Concept Learning and Semantic Parsing from Natural Language Explanations. Shashank Srivastava, Igor Labutov, Tom Mitchell. EMNLP. 2017
3. [RE:P3] AMR Parsing as Graph Prediction with Latent Alignment. Chunchuan Lyu, Ivan Titov · ACL · 2018

Meeting 6 (10.11.2019)

Embedding Methods: Points, Gaussian, Cones, Boxes, Cones, Hyperbolic-space methods [Emb]

Assigned Reading

1. [Emb:P1] Hyperbolic Entailment Cones for Learning Hierarchical Embeddings. Octavian-Eugen Ganea, Gary Bécigneul, Thomas Hofmann · ICML · 2018
2. [Emb:P2] Generalizing Point Embeddings using the Wasserstein Space of Elliptical Distributions. Boris Muzellec, Marco Cuturi · NeurIPS · 2018
3. [Emb:P3] Word Representations via Gaussian Embedding Luke Vilnis, Andrew McCallum · ICLR · 2015

Scribes:

• To be determined

Meeting 7 (10.18.2019)

Midpoint Project Presentations

1. Do We Really Need Knowledge Bases?
2. Effective and Efficient Entity Alias Predicting for Biomedical Text
3. Biomedical Entity Linking
4. Infusing Fairness in MINERVA Predictions
5. Commonsense Injected Text Encoding for Relation Extraction
6. Biomedical Mention Segmentation
7. SEM4Ent: The Simple Embedding Method for Entities
8. Knowledge Base Completion using Box Embedding
9. Learning Node Representations from Multiple Social Contexts.
10. Multi-Facet Embedding for Unsupervised Relation Extraction

Meeting 8 (10.25.2019)

Midpoint Project Presentations

1. Supervised Hierarchical Clustering with DeepSets
2. Evidence Selection in Long Documents
3. Reasoning Over Multiple Paths
4. Learning Causal Event Extraction Models from Disjoint Dataset
5. Efficient and online nearest neighbor search
6. Learning to understand scientific experiments
7. Boxes for Taxonomy Alignment
8. Learning Unified KG-Text Representations
9. Biomedical Within-Document Coreference
10. Scientific Paper Task and Problem Discovery, Contribution Identification and Tracking
11. Semi-supervised DIORA for NER

Meeting 9 (11.01.2019)

Lecture on: Graph Neural Networks & Learned Index Structures [GNN+Index] [Slides]

Lecture on: Question Answering, Reasoning, & Pretrained language models [QA]

No student presentations

Meeting 10 (11.08.2019)

Lecture on: Fairness & Adversarial Attacks [Fair]

Student Presentations on Question Answering, Reasoning, & Pretrained language models [QA]

Assigned Reading

• [QA:P1] Learning to Compose Neural Networks for Question Answering. Jacob Andreas, Marcus Rohrback, Trevor Darrell and Dan Klein. , NAACL 2016
• [QA:P2] End-to-End Differentiable Proving. Tim Rocktäschel, Sebastian Riedel, Neurips 2017
• [QA:P3] Language Models as Knowledge Bases. Fabio Petroni, Tim Rocktaschel, Patrick Lewis, Anton Bakhtin, Yuxiang Wu. Alexander H. Miller, Sebastian Riedel. EMNLP 2019

Meeting 11 (11.15.2019)

Student Presentations on: Graph Neural Networks & Learned Index Structures [GNN+Index]

Assigned Reading

• [GNN+Index:P1] The Case for Learned Index Structures. Tim Kraska, Alex Beutel, Ed Huai-hsin Chi, Jeffrey Dean, Neoklis Polyzotis in SIGMOD Conference 2017
• [GNN+Index:P2] Graph convolution over pruned dependency trees improves relation extraction. Yuhao Zhang, Peng Qi, Christopher D. Manning. EMNLP. 2018 and Attention Guided Graph Convolutional Networks for Relation Extraction. Zhijiang Guo, Yan Zhang, Wei Lu. ACL 2019.
• [GNN+Index:P3] Learning to Route in Similarity Graphs. Dmitry Baranchuk · Dmitry Persiyanov · Anton Sinitsin · Artem Babenko. ICML 2019

Student Presentations on: Fairness & Adversarial Attacks [Fair]

Assigned Reading

• [Fair:P1] Compositional Fairness Constraints for Graph Embeddings. Avishek Joey Bose, William L. Hamilton ICML 2019
• [Fair:P2] Fairness in Relational Domains. Golnoosh Farnadi, Behrouz Babaki, Lise Getoor AIES 2018
• [Fair:P3] Mitigating Gender Bias in Natural Language Processing: Literature Review. Tony Sun, Andrew Gaut, Shirlyn Tang, Yuxin Huang, Mai ElSherief, Jieyu Zhao, Diba Mirza, Elizabeth M. Belding-Royer, Kai-Wei Chang, William Yang Wang ACL 2019

Meeting 12 (11.22.2019)

Guest Lecture - Marco Serafini

Meeting 13 (12.06.2019)

Final Project Presentations