Deep Active Learning for Named Entity Recognition

By Yanyao Shen, Hyokun Yun, Zachary Chase Lipton, Yakov Kronrod, Anima Anandkumar. 2017

482 citations recorded by the source corpus.

Deep neural networks have advanced the state of the art in named entity recognition. However, under typical training procedures, advantages over classical methods emerge only with large datasets. As a result, deep learning is employed only when large public datasets or a large budget for manually labeling data is available. In this work, we show otherwise: by combining deep learning with active learning, we can outperform classical methods even with a significantly smaller amount of training data.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Incremental Learning from Scratch for Task-Oriented Dialogue Systems (2019, background): “It is worth noting that tricks such as mixing d t with the instances in the data pool and updating IDS for a small number of epochs (Shen et al., 2017) can be easily adopted to increase the utilization of labeled data.”
Assessing The Factual Accuracy of Generated Text (2019, background): “This is a very typical problem faced by any work using distant supervision, and is combated with methods like active learning [31].”
Hierarchical Contextualized Representation for Named Entity Recognition (2019, result): “We compare our model with the previous model that also reported results on it (Chiu and Nichols 2016; Shen et al. 2018; Ghaddar and Langlais 2018).”
Deep Learning on a Data Diet: Finding Important Examples Early in Training (2021, background): “Here, the goal is to select a small set of examples to label at any given iteration of training (see, e.g., [26–30], and references therein).”
Selection Via Proxy: Efficient Data Selection For Deep Learning (2019, methodology): “Like recent work for deep active learning (Shen et al., 2017; Sener & Savarese, 2018; Kirsch et al., 2019), we considered a batch setting with K rounds where we selected b K points in every round.”
BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning (2019, background): “There are a number of intuitive choices, such as model uncertainty and mutual information, and, in this paper, we focus on BALD [19], which has proven itself in the context of deep learning [13, 30, 20].”
A Survey on Deep Learning for Named Entity Recognition (2018, methodology): “A few studies [86]–[88], [94], [130] have explored RNN to decode tags.”
LEAN-LIFE: A Label-Efficient Annotation Framework Towards Learning from Explanation (2020, background): “Its model training follows an active learning strategy (Shen et al., 2017), which is shown to be a label-efficient, thus it attempts to minimize human annotation efforts.”
Detecting Independent Pronoun Bias with Partially-Synthetic Data Generation (2020, methodology): “For Humanin-the-Loop methods, like this paper, it known as Deep Active Bayesian Learning (Shen et al., 2018; Siddhant and Lipton, 2018).”
FedNER: Privacy-preserving Medical Named Entity Recognition with Federated Learning (2020, methodology): “, 2016), a variant of LSTM-CRF that uses a stacked Bi-LSTM for word representation; (6) CNN-CLSTM-CRF (Shen et al., 2017), a variant of CNN-LSTM-CRF, which uses a combination of CNN and LSTM to learn word representations; (7) ELMoNER, the medical NER model introduced in Section 3.”
FedNER: Medical Named Entity Recognition with Federated Learning (2020, methodology): “…representations; (5) S-LSTM-CRF (Lample et al., 2016), a variant of LSTM-CRF that uses a stacked Bi-LSTM for word representation; (6) CNN-CLSTM-CRF (Shen et al., 2017), a variant of CNN-LSTM-CRF, which uses a combination of CNN and LSTM to learn word representations; (7) ELMoNER, the medical NER…”
AlpacaTag: An Active Learning-based Crowd Annotation Framework for Sequence Tagging (2019, methodology): “For AlpacaTag , we apply an improved version named Maximum Normalized Log-Probability (MNLP) (Shen et al., 2018), which eliminates the effect of length of sequences by averaging: Simply put, we manage to utilize the current back-end model for measuring the informative-ness of unlabeled sentences…”
A Comprehensive Survey on Automatic Knowledge Graph Construction (2023, background): “[138] incrementally chose the most samples for NER labeling during the training procedures to mitigate the reliance on tagged samples.”
A Survey of Deep Active Learning (2020, methodology): “[34, 197, 199] uses the combination of DL and AL to determine how the technical level of NER (Named Entity Recognition) can be improved in the case of a small training set.”
SeqMix: Augmenting Active Sequence Labeling via Sequence Mixup (2020, background): “…methods (Clark et al., 2018; Chen et al., 2020b), ex-ternal weak supervision (Lison et al., 2020; Liang et al., 2020; Ren et al., 2020; Zhang et al., 2019; Yu et al., 2020) and active learning (Shen et al., 2017; Hazra et al., 2019; Liu et al., 2018; Fang et al., 2017; Gao et al., 2019).”