• Publications
  • Influence
KagNet: Knowledge-Aware Graph Networks for Commonsense Reasoning
TLDR
We propose a knowledge-aware reasoning framework for answering commonsense questions, which effectively utilizes external, structured commonsense knowledge graphs to perform explainable inferences. Expand
  • 74
  • 10
  • PDF
Multi-channel BiLSTM-CRF Model for Emerging Named Entity Recognition in Social Media
TLDR
In this paper, we present our multi-channel neural architecture for recognizing emerging named entity in social media messages, which we applied in the Novel and Emerging Named Entity Recognition shared task at the EMNLP 2017 Workshop on Noisy User-generated Text. Expand
  • 52
  • 3
  • PDF
Automatic Extraction of Commonsense LocatedNear Knowledge
TLDR
This paper aims to automatically extract the commonsense LOCATEDNEAR relation between physical objects from textual corpora. Expand
  • 12
  • 1
  • PDF
Mining Cross-Cultural Differences and Similarities in Social Media
TLDR
We present a lightweight yet effective approach, and evaluate it on two novel tasks: 1) mining cross-cultural differences of named entities and 2) finding similar terms for slang across languages. Expand
  • 11
  • 1
  • PDF
CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning
TLDR
In this paper, we present a constrained text generation task, CommonGen associated with a benchmark dataset, to explicitly test machines for the ability of generative commonsense reasoning. Expand
  • 6
  • 1
  • PDF
Can BERT Reason? Logically Equivalent Probes for Evaluating the Inference Capabilities of Language Models
TLDR
We present a procedure that allows for the systematized probing of both PTLMs' inference abilities and robustness. Expand
  • 1
  • 1
Neural Adaptation Layers for Cross-domain Named Entity Recognition
TLDR
In this paper, we empirically investigate effective methods for conveniently adapting an existing, well-trained neural NER model for a new domain. Expand
  • 30
  • PDF
NERO: A Neural Rule Grounding Framework for Label-Efficient Relation Extraction
TLDR
Deep neural models for relation extraction tend to be less reliable when perfectly labeled data is limited. Expand
  • 10
  • PDF
AlpacaTag: An Active Learning-based Crowd Annotation Framework for Sequence Tagging
TLDR
We introduce an open-source web-based data annotation framework (AlpacaTag) for sequence tagging tasks such as named-entity recognition (NER) and sequence labeling with recommendations powered by active learning. Expand
  • 8
  • PDF
TriggerNER: Learning with Entity Triggers as Explanations for Named Entity Recognition
TLDR
In this paper, we introduce "entity triggers," an effective proxy of human explanations for facilitating label-efficient learning of NER models. Expand
  • 6
  • PDF