Abstract:Extracting entities and relations for types of interest from text is important for understanding massive text corpora. Traditionally, systems of entity relation extraction have relied on human-annotated corpora for training and adopted an incremental pipeline. Such systems require additional human expertise to be ported to a new domain, and are vulnerable to errors cascading down the pipeline. In this paper, we investigate joint extraction of typed entities and relations with labeled data heuristically obtained from knowledge bases (i.e., distant supervision). As our algorithm for type labeling via distant supervision is context-agnostic, noisy training data poses unique challenges for the task. We propose a novel domain-independent framework, called CoType, that runs a data-driven text segmentation algorithm to extract entity mentions, and jointly embeds entity mentions, relation mentions, text features and type labels into two low-dimensional spaces (for entity and relation mentions respectively), where, in each space, objects whose types are close will also have similar representations. CoType, then using these learned embeddings, estimates the types of test (unlinkable) mentions. We formulate a joint optimization problem to learn embeddings from text corpora and knowledge bases, adopting a novel partial-label loss function for noisy labeled data and introducing an object "translation" function to capture the cross-constraints of entities and relations on each other. Experiments on three public datasets demonstrate the effectiveness of CoType across different domains (e.g., news, biomedical), with an average of 25% improvement in F1 score compared to the next best method.

METIC: Multi-Instance Entity Typing from Corpus.

Boosting Collective Entity Linking via Type-Guided Semantic Embedding.

Knowledge Representation Learning with Entity Descriptions, Hierarchical Types, and Textual Relations

Attributed and Predictive Entity Embedding for Fine-Grained Entity Typing in Knowledge Bases.

Multi-view Contrastive Learning for Entity Typing over Knowledge Graphs

CoType: Joint Extraction of Typed Entities and Relations with Knowledge Bases

Cross-Lingual Type Inference.

Type-enriched Hierarchical Contrastive Strategy for Fine-Grained Entity Typing

Transformer-based Entity Typing in Knowledge Graphs

The Integration of Semantic and Structural Knowledge in Knowledge Graph Entity Typing

Connecting Embeddings Based on Multiplex Relational Graph Attention Networks for Knowledge Graph Entity Typing

Calibrated Seq2seq Models for Efficient and Generalizable Ultra-fine Entity Typing

A Chinese Corpus for Fine-grained Entity Typing

KCAT: A Knowledge-Constraint Typing Annotation Tool

Fine-Grained Entity Typing for Domain Independent Entity Linking

Liberal Entity Extraction: Rapid Construction of Fine-Grained Entity Typing Systems

Put It Back: Entity Typing with Language Model Enhancement

Knowledge Graph Entity Typing Via Learning Connecting Embeddings

Multimodal Entity Tagging with Multimodal Knowledge Base

Entity Identification as Multitasking

Type Hierarchy Enhanced Heterogeneous Network Embedding for Fine-Grained Entity Typing in Knowledge Bases