Papers with neural sequence labeling

4 papers
NCRF++: An Open-source Neural Sequence Labeling Toolkit (P18-4)

Copied to clipboard

Challenge: Existing statistical approaches to neural sequence labeling have been used for many tasks.
Approach: They describe a toolkit for neural sequence labeling that provides a CRF inference layer for quick implementation.
Outcome: The toolkit is based on PyTorch and can be run on GPUs.
Hybrid semi-Markov CRF for Neural Sequence Labeling (P18-2)

Copied to clipboard

Challenge: Existing conditional random fields (CRFs) use hand-crafted features to perform sequence labeling tasks.
Approach: They propose to use semi-Markov conditional random fields for neural sequence labeling in natural language processing to extract features from segments instead of words.
Outcome: The proposed model achieves state-of-the-art when no external knowledge is used.
Hierarchically-Refined Label Attention Network for Sequence Labeling (D19-1)

Copied to clipboard

Challenge: Conditional random fields (CRF) is a powerful model for statistical sequence labeling, but it does not give much information gain over strong neural encoding.
Approach: They propose a hierarchically-refined label attention network which captures potential long-term label dependency by giving each word incrementally refined label distributions with hierarchical attention.
Outcome: The proposed model improves POS tagging accuracy and speeds up training and testing compared to the current model.
Handling Rare Entities for Neural Sequence Labeling (2020.acl-main)

Copied to clipboard

Challenge: Recent approaches to sequence labeling have been based on statistical models but a challenge is from the data sparsity problem.
Approach: They propose to use local context reconstruction to implicitly incorporate contextual information into their representations.
Outcome: The proposed model outperforms all previous methods on multiple benchmark datasets and achieves new start-of-the-art results.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations