Skip to content

Latest commit

 

History

History
115 lines (84 loc) · 6.16 KB

aaai20.md

File metadata and controls

115 lines (84 loc) · 6.16 KB

Knowledge Graphs at AAAI 2020

KRL

Rule-Guided Compositional Representation Learning on Knowledge Graphs AAAI 2020. Niu et al. [Paper]

Triple2Vec: Learning Triple Embeddings from Knowledge Graphs. AAAI 2020. Fionda et al. [Paper]

InteractE: Improving Convolution-based Knowledge Graph Embeddings by Increasing Feature Interactions AAAI 2020. Vashishth et al. [Paper] [Code]

KGC

Few-Shot Knowledge Graph Completion. AAAI 2020. Zhang et al. [Paper]

Learning Hierarchy-Aware Knowledge Graph Embeddings for Link Prediction AAAI 2020. Zhang et al. [Paper]

HAKE, Hierarchy-Aware Knowledge Embedding; capture semantic hierarchies; map entities into the polar coordinate system; radial coordinate for representing levels of the hierarchy and angular coordinate for distinguishing entities at the same level of hierarchy

Commonsense Knowledge Base Completion with Structural and Semantic Context AAAI 2020. Malaviya et al.. [Paper]

Understanding the semantic content of sparse word embeddings using a commonsense knowledge base AAAI 2020 Workshop. [Paper]

Relational Graph Neural Network with Hierarchical Attention for Knowledge Graph Completion AAAI 2020. Zhang et al.. [Paper]

encoder-decoder framework, hierarchically attentive graph encoder

Contextual Parameter Generation for Knowledge Graph Link Prediction. AAAI 2020. Stoica et al.. [Paper] [Paper]

Diachronic Embedding for Temporal Knowledge Graph Completion. AAAI 2020. Goel et al.. [Paper] [Code]

Relation Extraction

Are Noisy Sentences Useless for Distant Supervised Relation Extraction? AAAI 2020. Shang et al.. [Paper]

Self-Attention Enhanced Selective Gate with Entity-Aware Embedding for Distantly Supervised Relation Extraction AAAI 2020. Li et al. [Paper]

Entity Alignment

Knowledge Graph Alignment Network with Gated Multi-hop Neighborhood Aggregation AAAI 2020. Zequn Sun, Chengming Wang, Wei Hu1, Muhao Chen, Jian Dai, Wei Zhang, Yuzhong Qu [Paper] [Code]

AliNet to mitigate the non-isomorphism of neighborhood structures
graph neural networks, attention, a relation loss to refine entity representations

Coordinated Reasoning for Cross-Lingual Knowledge Graph Alignment. AAAI 2020. Xu et al.. [Paper]

Entity Linking

LATTE: Latent Type Modeling for Biomedical Entity Linking AAAI 2020. Zhu et al.. [Paper]

Improving Entity Linking by Modeling Latent Entity Type Information AAAI 2020. Chen et al. [Paper]

Named Entity Recognition

Leveraging Multi-token Entities in Document-level Named Entity Recognition AAAI 2020. Hu et al. [Paper]

Robust Named Entity Recognition with Truecasing Pretraining. AAAI 2020. Mayhew et al. [Paper]

Meta Relational Learning

Generative Adversarial Zero-Shot Relational Learning for Knowledge Graphs AAAI 2020. Qin et al. [Paper]

Natural Language Understanding

Evaluating Commonsense in Pre-trained Language Models AAAI 2020. Zhou et al. [Paper] [Code & Dataset]

study the commonsense ability of GPT, BERT, XLNet, and RoBERTa test the robustness of models by making dual test cases

Latent Relation Language Models. AAAI 2020. Hayashi et al. [Paper]

LRLM parameterizes the joint distribution over the words in a document annotate the posterior probability of entity spans for a given text through relations

K-BERT: enabling language representation with knowledge graph AAAI 2020. Liu et al. [Paper] [Code]

K-BERT, Knowledge-enabled Bidirectional Encoder Representation from Transformers Method: injecting triples into sentences, soft-position and visible matrix to overcome knowledge noise
Task: Chinese NLP, open-domain and specific-domain tasks

Differentiable Reasoning on Large Knowledge Bases and Natural Language AAAI 2020. Minervini et al. [Paper]

GNTP, Greedy Neural Theorem Provers
Idea: extend NTPs to address their complexity and scalability limitation, reducing the number of candidate proof paths and introducing an attention mechanism for rule induction
Task: link prediction

Reasoning on Knowledge Graphs using Debate Dynamics AAAI 2020. Hildebrandt et al. [Paper] [Code]

R2D2 (Reveal Relations using Debate Dynamics)
Task: triple classification
Idea: triple classification as a debate game between two RL agents extracting relational paths as arguments
Method: reinforcement learning, a binary classifier judges the factual triples as thesis or antithesis

Knowledge Graph Grounded Goal Planning for Open-Domain Conversation Generation AAAI 2020. Xu et al.. [Paper]

Towards Scalable Multi-Domain Conversational Agents:The Schema-Guided Dialogue Dataset. AAAI 2020. Rastogi et al. [Paper]

Computer Vision

Knowledge Graph Transfer Network for Few-Shot Recognition AAAI 2020. Chen et al.. [Paper]