Pith. sign in

REVIEW 4 major objections 4 minor 40 references

NE-LP: Normalized Entropy and Loss Prediction based Sampling for Active Learning in Chinese Word Segmentation on EHRs

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Active learning that scores sentences by normalized entropy plus a predicted segmentation loss selects more informative electronic health record sentences than standard uncertainty sampling, as measured by F1 over active-learning rounds.

desk verdict A reasonable active-learning recipe for clinical Chinese word segmentation, but the claimed gains over uncertainty baselines aren't established because the baselines look mis-specified and the key ablation is missing. read the letter →

arxiv 1908.08419 v3 pith:5A5C4IFX submitted 2019-08-22 cs.CL

Tingting Cai , Zhiyuan Ma , Hong Zheng , Yangming Zhou This is my paper · ORCID
classification cs.CL
keywords activelearningChinesewordsegmentationelectronichealthrecordslosspredictionnormalizedentropyBiLSTM-CRFself-attentionbigramfeatures
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Chinese electronic health records contain no word delimiters, so clinical text needs segmentation models trained on labeled data, but medical annotation is slow because it requires expert annotators. This paper proposes an active-learning sampling strategy, NE-LP, that scores unlabeled sentences by a weighted sum of two signals: normalized token entropy from a BiLSTM-CRF segmenter and a predicted segmentation loss from a self-attention head trained jointly with the segmenter. On a cardiovascular EHR corpus, the authors report that selecting sentences by this score consistently yields higher F1 than least confidence, maximum token entropy, minimum token margin, or random selection across active-learning rounds. The intended payoff is the same segmentation quality from fewer expert-annotated sentences, which matters for specialized clinical domains where large labeled corpora are unavailable.

What carries the argument

The load-bearing object is a joint model: a BiLSTM-CRF word segmenter with an attached self-attention loss-prediction head. The two share the BiLSTM encoder, and the joint objective is $L_{joint}=L_{Seg}+\lambda L_{Loss}$, where $L_{Loss}$ is the mean squared error between the predicted loss and the true segmentation loss of the segmenter. Inputs are character embeddings concatenated with pretrained bigram embeddings, which capture interactions between adjacent characters; on this corpus bigram features outperform trigram and four-gram because two-character words dominate the data. During active learning, the same joint model computes both the normalized entropy needed for the uncertainty term and the predicted loss, so no separate selection model needs to be trained. The NE-LP score in Equation (16) is the mechanism that turns raw model outputs into a ranking of sentences for annotation.

What would settle it

Fix the segmenter and replace the loss-prediction term with a control trained on randomly shuffled segmentation losses; if NE-LP's F1 advantage over normalized-entropy-only sampling does not disappear or shrink sharply, then the reported gain is not caused by loss prediction. A direct check is to plot, on held-out sentences, the correlation between predicted loss and the segmenter's actual loss after retraining, and see whether the top-ranked sentences by NE-LP actually produce the largest F1 gains.

Watch

Extended reading notes

Core claim

The central claim is that the NE-LP score, $S_{NE-LP}(x)=\alpha\,Uncertainty(x)+\beta\,Loss(x)$, identifies more useful sentences for the next labeling round than uncertainty alone. Uncertainty is the entropy of the segmenter's per-character tag distribution over the BMES labels, normalized by $\log(1/N)$ and scaled by $1/\sqrt{Len}$; loss is the output of a self-attention layer trained to reproduce the segmenter's actual segmentation loss on labeled data. The authors demonstrate the claim on an EHR dataset by comparing learning curves over ten active-learning rounds at two initial-label ratios, 3:7 and 1:9. In every reported round and both ratios, NE-LP reaches a higher testing-set F1 than LC, MTE, MTM, and RAND, with the advantage over MTE larger when the initial labeled set is larger.

Load-bearing premise

The whole gain rests on the assumption that a loss predictor trained on the current labeled set can judge which unlabeled sentences will be hardest for the segmenter; if predicted loss is a poor proxy on unlabeled data, NE-LP degenerates to normalized entropy sampling.

Editorial extensions

If this is right

  • If NE-LP is correct, fewer expert-labeled EHR sentences are needed to reach a target segmentation F1, lowering annotation cost in medical NLP.
  • The method remains ahead of uncertainty baselines when the initial labeled set is small (1:9 ratio), suggesting it helps in especially low-resource settings.
  • Bigram features improve the segmenter on this corpus, and NE-LP benefits from them; the gain is explained by the prevalence of two-character words in clinical text.
  • The loss-prediction component contributes beyond normalized entropy, since setting both weights to 1 outperforms weighting either signal alone, and the authors state the method is intended to transfer to other sequence labeling tasks such as named entity recognition and relation extraction.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An untested extension: the loss predictor's marginal value may shrink as the seed set shrinks, because the reported gap between NE-LP and MTE narrows under the 1:9 split; measuring the gap across several seed sizes would show whether loss prediction is most useful once a moderate labeled base exists.
  • The $1/\sqrt{Len}$ normalization is an ad hoc scaling, so a natural follow-up is to compare NE-LP with per-length normalization or learned combination weights to see whether sentence length distorts the entropy-loss sum.
  • Because the loss predictor is trained on labels also used to train the segmenter, its usefulness on unlabeled sentences depends on extrapolation; long sentences, rare medical terms, and out-of-vocabulary drugs would stress the predictor and could reveal whether the reported gain persists outside the original EHR distribution.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes NE-LP, an active learning sampling strategy for Chinese Word Segmentation (CWS) on electronic health records. The method combines a normalized entropy term, intended to measure segmentation uncertainty, with a loss prediction term from a self-attention submodel attached to a BiLSTM-CRF segmenter. The joint model is trained on labeled data, and at each active learning round the highest-scoring unlabeled sentences are selected for annotation. Experiments are conducted on cardiovascular EHRs from Shuguang Hospital. The paper reports that NE-LP consistently outperforms least confidence (LC), maximum token entropy (MTE), minimum token margin (MTM), and random sampling in F1-score, under two initial labeled set ratios and with or without bigram features.

Significance. If the central claim were fully supported, the paper would make a useful contribution: active learning for domain-specific CWS is a practical problem, the idea of combining uncertainty with a learned loss prediction model is plausible, and the use of real Chinese EHR data with an explicit annotation bottleneck is a strength. The paper also provides useful auxiliary comparisons, including several open-source CWS tools and multiple neural segmenters, and it clearly quantifies the gains from bigram features. However, the main comparative evidence is undermined by non-standard baseline definitions in the text, a mismatch between those definitions and the reported learning curves, and the absence of statistical repeatability measures. The work is therefore better framed as a proposal with preliminary evidence than as an established improvement over conventional uncertainty sampling.

major comments (4)
  1. [Section 2.2, Eqs. (2)-(3)] The definitions of MTE and MTM are not standard token-level uncertainty scores. In Eq. (2), the summand does not depend on i, so S_MTE(x) reduces to -N * p(y*|x) log p(y*|x), which is a function of only the best-path probability. In Eq. (3), the margin is computed between sequence-level path probabilities rather than between the top two tag probabilities at each token position, which is the conventional margin for sequence labeling as used in Marcheggiani and Artieres (2014), reference [23]. Because -N p log p is strictly decreasing in p for p in (0,1), Eq. (2) should rank sentences nearly identically to LC in Eq. (1). The learning curves in Figs. 5 and 6 show noticeably different behavior for LC and MTE, which indicates that the implemented baselines are not the ones defined in Eqs. (1)-(3). Without the code or an exact protocol, the reader cannot verify that NE-LP is compared against genuine mainstream uncertainty strategies rather than weakened versions. This directly affects the abstract's claim of consistent outperformance and should be fixed by correcting the baseline definitions and re-running, or by clearly describing the actual implementations used.
  2. [Section 4.3.4, Eq. (16)] The weighting coefficients alpha and beta are selected by comparing learning curves in Fig. 4 on the same test set used in the main comparisons, and the paper then fixes alpha = beta = 1 for all subsequent experiments. This is a form of test-set selection that can inflate the reported gains of NE-LP over the baselines, because the hyperparameters are chosen to maximize the displayed F1 advantage. A held-out validation set or nested selection procedure should be used, and the sensitivity of the main comparison to alpha and beta should be reported. Without this, the reported margins in Figs. 5 and 6 are not a fair estimate of expected performance.
  3. [Section 4.3.5, Figs. 5-6] The central experimental evidence consists of single learning curves with no error bars, no multiple runs, and no significance tests. The reported gaps between NE-LP and MTE are small in several iterations, especially in the 1:9 setting of Fig. 6, where the paper itself notes that the curves are close. A single run cannot support the claim that NE-LP 'always' or 'consistently' outperforms MTE. The paper should report means and standard deviations over multiple random seeds, and where appropriate paired significance tests. Additionally, no NE-only ablation (beta = 0) is reported, so the specific contribution of the loss prediction term to the claimed gains is not demonstrated. Section 3.4 states that the informativeness assumption is verified in the experiments, but the experiments only compare the combined strategy against baselines, not NE-LP against NE alone.
  4. [Section 3.4, Eq. (15)] The sampling score in Eq. (15) is called normalized entropy, but the notation is ambiguous: pSeg(x) appears inside a sum over N classes without a token index, and the denominator contains log(1/N) * sqrt(Len), where Len is the sequence length. If pSeg(x) is the whole-sequence probability of the best tag sequence, then Eq. (15) is not a token-level normalized entropy and the denominator's dependence on sequence length is unmotivated. If pSeg(x) is instead the marginal probability of a tag at a token, then the formula needs a sum over tokens and a clear definition of how token entropies are aggregated into a sentence score. The authors should clarify the exact quantity used to rank sentences, since this is the core of the proposed method.
minor comments (4)
  1. [Abstract] The abstract contains a typo: 'Specically' should be 'Specifically'.
  2. [Algorithm 1] Line 5 reads 'label U by fτ', which is unclear for an active learning setup where U is the unlabeled pool; it should say 'predict labels for U with fτ' or 'obtain current predictions on U'.
  3. [Section 2.2, Eq. (3)] The prose says MTM 'subtracts the highest probability by the lowest one'; this should read 'subtracts the second-highest probability from the highest one' to match standard margin sampling and the notation max'.
  4. [References] Reference [31] contains a corrupted author field ('Kaiser, /suppress L.'); this should be corrected to the standard Vaswani et al. citation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: NE-LP is an empirical active-learning heuristic whose loss-prediction component and uncertainty term are not defined in terms of the reported outcome, and the comparisons against baselines provide independent evidence.

full rationale

The paper's central claim is that the NE-LP sampling strategy (Eq. 16) outperforms conventional uncertainty-based strategies on an EHR Chinese word segmentation task. The derivation chain is not circular: normalized entropy (Eq. 15) is a standard uncertainty measure computed from segmenter posterior probabilities, and the loss prediction term is an auxiliary model trained with Eq. (14) to regress the current segmenter's real segmentation loss on labeled data, following the external method of Yoo and Kweon [38]. Applying that trained predictor to unlabeled sentences is a genuine inductive step, not a quantity that is equal to the reported F1 by construction. The hyperparameters alpha, beta, and lambda are selected from small grids and are tuning choices, not fitted constants that force the final comparison. Citations with author overlap ([17], [29], [33]) support only background statements about RNNs, BiLSTM, and CRF and are not load-bearing for the novelty claim. The paper's real weaknesses are benchmark-related rather than circular: the MTE and MTM baseline equations in Eqs. (2)-(3) are nonstandard and the paper does not report an ablation with beta=0 to isolate the loss-prediction contribution. Those concerns affect validity of the empirical superiority claim, but they do not amount to a derivation that reduces to its own inputs. No circular step can be exhibited from the paper's equations, so the appropriate verdict is no significant circularity.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim relies on a small set of hand-tuned coefficients and two assumptions about loss prediction and informativeness. There are no new physical or conceptual entities.

free parameters (4)
  • alpha (weight of normalized entropy in NE-LP) = 1
    Chosen after comparing five weight settings in Fig. 4 on the same dataset; this tuning is not held out from the final comparison.
  • beta (weight of loss prediction in NE-LP) = 1
    Same experiment as alpha; optimal on the dataset used for evaluation.
  • lambda (weight of loss prediction loss in joint loss) = 1
    Set empirically in Section 3.3.2 without a sensitivity study.
  • active learning iterations M = 10
    Fixed because 'each sampling strategy does not improve obviously after 10 iterations', a post hoc choice on the same data.
assumptions (3)
  • ad hoc to paper Samples with both high uncertainty and high predicted loss are the most informative for the current segmenter.
    Stated as a hypothesis in Section 3.4 and verified only on the private EHR dataset; the whole NE-LP strategy depends on it.
  • domain assumption The loss prediction model, trained on labeled data, generalizes to unlabeled sentences so its predicted loss is a meaningful selection signal.
    Implied by the method in Section 3.3; no evidence is given that predicted loss transfers from the labeled distribution to unlabeled examples.
  • domain assumption BiLSTM-CRF with bigram embeddings is an appropriate base segmenter for this task.
    Justified by model comparisons in Table 5, but the choice is on the same data and the margin over CNN-BiLSTM-CRF is 0.08%.

how reviews work

0 comments
Cite this review

Pith. "Pith review of NE-LP: Normalized Entropy and Loss Prediction based Sampling for Active Learning in Chinese Word Segmentation on EHRs." pith.science (2026). https://pith.science/paper/5A5C4IFX

@misc{pith2026190808419,
  author       = {Pith},
  title        = {Pith review of: NE-LP: Normalized Entropy and Loss Prediction based Sampling for Active Learning in Chinese Word Segmentation on EHRs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5A5C4IFX}},
  note         = {Machine review of arXiv:1908.08419}
}
read the original abstract

Electronic Health Records (EHRs) in hospital information systems contain patients' diagnosis and treatments, so EHRs are essential to clinical data mining. Of all the tasks in the mining process, Chinese Word Segmentation (CWS) is a fundamental and important one, and most state-of-the-art methods greatly rely on large-scale of manually-annotated data. Since annotation is time-consuming and expensive, efforts have been devoted to techniques, such as active learning, to locate the most informative samples for modeling. In this paper, we follow the trend and present an active learning method for CWS in EHRs. Specically, a new sampling strategy combining Normalized Entropy with Loss Prediction (NE-LP) is proposed to select the most representative data. Meanwhile, to minimize the computational cost of learning, we propose a joint model including a word segmenter and a loss prediction model. Furthermore, to capture interactions between adjacent characters, bigram features are also applied in the joint model. To illustrate the effectiveness of NE-LP, we conducted experiments on EHRs collected from the Shuguang Hospital Affiliated to Shanghai University of Traditional Chinese Medicine. The results demonstrate that NE-LP consistently outperforms conventional uncertainty-based sampling strategies for active learning in CWS.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 37 canonical work pages

  1. [23]

    In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, pp

    Marcheggiani, D., Artieres, T.: An experimental com- parison of active learning strategies for partially labele d sequences. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, pp. 898–906 (2014)

    work page 2014

  2. [1]

    Machine Learning 2(4), 319–342 (1988)

    Angluin, D.: Queries and concept learning. Machine Learning 2(4), 319–342 (1988)

    work page 1988

  3. [2]

    IEEE Transactions on Neural Networks 5(2), 157–166 (1994)

    Bengio, Y., Simard, P., Frasconi, P., et al.: Learning long-term dependencies with gradient descent is difficult. IEEE Transactions on Neural Networks 5(2), 157–166 (1994)

    work page 1994

  4. [3]

    In: Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pp

    Chen, X., Qiu, X., Zhu, C., Liu, P., Huang, X.: Long short-term memory neural networks for Chinese word segmentation. In: Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pp. 1197–1206 (2015)

    work page 2015

  5. [4]

    IEEE Access 8, 16387–16396 (2020)

    Cheng, K., Yue, Y., Song, Z.: Sentiment classification based on part-of-speech and self-attention mechanism. IEEE Access 8, 16387–16396 (2020)

    work page 2020

  6. [5]

    Journal of Machine Learning Research 12(Aug), 2493–2537 (2011)

    Collobert, R., Weston, J., Bottou, L., Karlen, M., Kavukcuoglu, K., Kuksa, P.: Natural language process- ing (almost) from scratch. Journal of Machine Learning Research 12(Aug), 2493–2537 (2011)

    work page 2011

  7. [6]

    Bioinformat- ics (Oxford, England) 14(9), 755–763 (1998)

    Eddy, S.R.: Profile hidden markov models. Bioinformat- ics (Oxford, England) 14(9), 755–763 (1998)

    work page 1998

  8. [7]

    Investigating Self-Attention Network for Chinese Word Segmentation

    Gan, L., Zhang, Y.: Investigating self-attention net- work for Chinese word segmentation. arXiv preprint arXiv:1907.11512 (2019)

    work page Pith review arXiv 2019

Show all 40 references
  1. [8]

    Procedia Computer Science 124, 544–551 (2017) 12 Tingting Cai et al

    Gesulga, J.M., Berjame, A., Moquiala, K.S., Galido, A.: Barriers to electronic health record system implementa- tion and information systems resources: A structured re- view. Procedia Computer Science 124, 544–551 (2017) 12 Tingting Cai et al

    2017

  2. [9]

    In: Advances in neural informa- tion processing systems, pp

    Gilad-Bachrach, R., Navot, A., Tishby, N.: Query by committee made real. In: Advances in neural informa- tion processing systems, pp. 443–450 (2006)

    2006

  3. [10]

    arXiv preprint arXiv:1402.3722 (2014)

    Goldberg, Y., Levy, O.: Word2Vec explained: deriv- ing mikolov et al.’s negative-sampling word-embedding method. arXiv preprint arXiv:1402.3722 (2014)

    2014 arXiv

  4. [11]

    arXiv preprint arXiv:1112.5745 (2011)

    Houlsby, N., Husz´ ar, F., Ghahramani, Z., Lengyel, M.: Bayesian active learning for classification and preference learning. arXiv preprint arXiv:1112.5745 (2011)

    2011 arXiv

  5. [12]

    In: International Conference on Learning Representations (2018)

    Hu, P., Lipton, Z.C., Anandkumar, A., Ramanan, D.: Active learning with partial feedback. In: International Conference on Learning Representations (2018)

    2018

  6. [13]

    Scientific Programming 2018, 1–10 (2018)

    Huang, H., Wang, H., Jin, D.: A low-cost named entity recognition research based on active learning. Scientific Programming 2018, 1–10 (2018)

    2018

  7. [14]

    arXiv preprint arXiv:1905.08454 (2019)

    Jiang, W., Tang, Y.: A seq-to-seq transformer premised temporal convolutional network for Chinese word seg- mentation. arXiv preprint arXiv:1905.08454 (2019)

    2019 arXiv

  8. [15]

    Neural Computing and Applications 32(1), 41–49 (2020)

    La Su, Y., Liu, W.: Research on the lstm mongolian and chinese machine translation based on morpheme encod- ing. Neural Computing and Applications 32(1), 41–49 (2020)

    2020

  9. [16]

    In: Proceedings of the Eighteenth International Conference on Machine Learning, pp

    Lafferty, J.D., McCallum, A., Pereira, F.C.: Conditiona l random fields: Probabilistic models for segmenting and labeling sequence data. In: Proceedings of the Eighteenth International Conference on Machine Learning, pp. 282– 289 (2001)

    2001

  10. [17]

    In: IEEE International Conference on Bioinformatics and Biomedicine, pp

    Lei, L., Zhou, Y., Zhai, J., Zhang, L., Fang, Z., He, P., Gao, J.: An effective patient representation learn- ing for time-series prediction tasks based on ehrs. In: IEEE International Conference on Bioinformatics and Biomedicine, pp. 885–892. IEEE (2018)

    2018

  11. [18]

    In: Proceedings of the 17th An- nual International Conference on Research and Develop- ment in Information Retrieval, pp

    Lewis, D.D., Gale, W.A.: A sequential algorithm for training text classifiers. In: Proceedings of the 17th An- nual International Conference on Research and Develop- ment in Information Retrieval, pp. 3–12. Springer (1994)

    1994

  12. [19]

    In: Proceedings of International Conference on Computational Linguistics 2012: Posters, pp

    Li, S., Zhou, G., Huang, C.R.: Active learning for Chi- nese word segmentation. In: Proceedings of International Conference on Computational Linguistics 2012: Posters, pp. 683–692 (2012)

    2012

  13. [20]

    Neurocomput- ing 338, 46–54 (2019)

    Liu, J., Wu, F., Wu, C., Huang, Y., Xie, X.: Neural chi- nese word segmentation with dictionary. Neurocomput- ing 338, 46–54 (2019)

    2019

  14. [21]

    arXiv preprint arXiv:2001.02524 (2020)

    Liu, M., Tu, Z., Wang, Z., Xu, X.: LTP: A new ac- tive learning strategy for Bert-CRF based named entity recognition. arXiv preprint arXiv:2001.02524 (2020)

    2020 arXiv

  15. [22]

    In: Proceedings of the 2018 Conference on Empirical Methods in Natural Lan- guage Processing, pp

    Ma, J., Ganchev, K., Weiss, D.: State-of-the-art Chines e word segmentation with bi-lstms. In: Proceedings of the 2018 Conference on Empirical Methods in Natural Lan- guage Processing, pp. 4902–4908 (2018)

    2018

  16. [24]

    In: Proceedings of the 20th international conference on Computational Linguistics, pp

    Peng, F., Feng, F., McCallum, A.: Chinese segmentation and new word detection using conditional random fields. In: Proceedings of the 20th international conference on Computational Linguistics, pp. 562–568. Association for Computational Linguistics (2004)

    2004

  17. [25]

    IEEE Access 7, 12993–13002 (2019)

    Shao, D., Zheng, N., Yang, Z., Chen, Z., Xiang, Y., Xian, Y., Yu, Z.: Domain-specific Chinese word segmentation based on bi-directional long-short term memory model. IEEE Access 7, 12993–13002 (2019)

    2019

  18. [26]

    In: CIPS-SIGHAN Joint Conference on Chinese Language Processing (2010)

    Song, H., Yao, T., Kit, C., Cai, D.: Active learning based corpus annotation. In: CIPS-SIGHAN Joint Conference on Chinese Language Processing (2010)

    2010

  19. [27]

    Journal of Machine Learn- ing Research 15(1), 1929–1958 (2014)

    Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.: Dropout: a simple way to prevent neu- ral networks from overfitting. Journal of Machine Learn- ing Research 15(1), 1929–1958 (2014)

    2014

  20. [28]

    Neural Computing and Applications (2020)

    Sun, D., Yaqot, A., Qiu, J., Rauchhaupt, L., Jumar, U., Wu, H.: Attention-based deep convolutional neural net- work for spectral efficiency optimization in mimo sys- tems. Neural Computing and Applications (2020)

    2020

  21. [29]

    arXiv preprint arXiv:2006.00464 (2020)

    Tang, P., Yang, P., Shi, Y., Zhou, Y., Lin, F., Wang, Y.: Recognizing Chinese judicial named entity using BiLSTM-CRF. arXiv preprint arXiv:2006.00464 (2020)

    2020 arXiv

  22. [30]

    Institution of Engineering and Technology Computer Vision 13(2), 194–205 (2018)

    Tang, X., Du, B., Huang, J., Wang, Z., Zhang, L.: On combining active and transfer learning for medical data classification. Institution of Engineering and Technology Computer Vision 13(2), 194–205 (2018)

    2018

  23. [31]

    In: Advances in Neural Information Processing Systems, pp

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, /suppress L., Polosukhin, I.: Atten- tion is all you need. In: Advances in Neural Information Processing Systems, pp. 5998–6008 (2017)

    2017

  24. [32]

    arXiv preprint arXiv:1711.04411 (2017)

    Wang, C., Xu, B.: Convolutional neural network with word embeddings for Chinese word segmentation. arXiv preprint arXiv:1711.04411 (2017)

    2017 arXiv

  25. [33]

    Journal of biomedical informatics 92, 103–133 (2019)

    Wang, Q., Zhou, Y., Ruan, T., Gao, D., Xia, Y., He, P.: Incorporating dictionaries into deep neural networks for the Chinese clinical named entity recognition. Journal of biomedical informatics 92, 103–133 (2019)

    2019

  26. [34]

    In: Proceedings of the 27th International Conference on Computational Linguistics, pp

    Xing, J., Zhu, K., Zhang, S.: Adaptive multi-task transfer learning for Chinese word segmentation in medical text. In: Proceedings of the 27th International Conference on Computational Linguistics, pp. 3619–3630 (2018)

    2018

  27. [35]

    In: Proceedings of the second SIGHAN workshop on Chinese language processing-Volume 17, pp

    Xue, N., Shen, L.: Chinese word segmentation as lmr tag- ging. In: Proceedings of the second SIGHAN workshop on Chinese language processing-Volume 17, pp. 176–179. Association for Computational Linguistics (2003)

    2003

  28. [36]

    In: National CCF Conference on Natural Language Pro- cessing and Chinese Computing, pp

    Yan, Q., Wang, L., Li, S., Liu, H., Zhou, G.: Active learning for Chinese word segmentation on judgements. In: National CCF Conference on Natural Language Pro- cessing and Chinese Computing, pp. 839–848. Springer (2017)

    2017

  29. [37]

    Acta Au- tomatica Sinica 40(8), 1537–1562 (2014)

    Yang, J., Yu, Q., Guan, Y., Jiang, Z.: An overview of research on electronic medical record oriented named en- tity recognition and entity relation extraction. Acta Au- tomatica Sinica 40(8), 1537–1562 (2014)

    2014

  30. [38]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp

    Yoo, D., Kweon, I.S.: Learning loss for active learning. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 93–102 (2019)

    2019

  31. [39]

    In: Proceedings of the 20th Pacific Asia Conference on Language, Information and Computation, pp

    Zhao, H., Huang, C.N., Li, M., Lu, B.L.: Effective tag set selection in Chinese word segmentation via condi- tional random field modeling. In: Proceedings of the 20th Pacific Asia Conference on Language, Information and Computation, pp. 87–94 (2006)

    2006

  32. [40]

    In: Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, pp

    Zheng, X., Chen, H., Xu, T.: Deep learning for Chinese word segmentation and pos tagging. In: Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, pp. 647–657 (2013)

    2013

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.