CBOWRA: A Representation Learning Approach for Medication Anomaly Detection

Ju Gao; Kai Wang; Liang Zhao; Shengping Liu; Yangming Zhou; Zhiyuan Ma

REVIEW 4 major objections 5 minor 29 references

CBOWRA: A Representation Learning Approach for Medication Anomaly Detection

T0 review · 4 major / 5 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 A word-embedding model flags wrong prescriptions by finding the drug least likely to fit the diagnosis list.

desk verdict The CBOWRA paper is a reasonable idea undermined by an in-sample evaluation and a gold standard that the authors themselves show flags correct prescriptions as errors. read the letter →

arxiv 1908.07147 v2 pith:KIAYSIMV submitted 2019-08-20 cs.CL cs.LG

Liang Zhao , Zhiyuan Ma , Yangming Zhou , Kai Wang , Shengping Liu , Ju Gao This is my paper · ORCID

classification cs.CLcs.LG

keywords medicationanomalydetectionelectronichealthrecordscontinuousbagofwordsrepresentationlearningrankingaccumulationsemanticconsistency

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

The paper tries to show that a simple representation learning method, built on the continuous bag-of-words idea, can detect medication anomalies in electronic health records without hand-crafted expert features. The key move is to treat a patient's diagnoses as context and each prescribed drug as a candidate central word, then flag the drug the model thinks is least likely to appear in that context. If the claim is right, hospitals could automatically rank suspicious prescriptions with top-5 accuracy above 0.84 on real records, saving clinicians from manual chart review. The method is tested against distance-based outlier detection, naive Bayes variants, and knowledge-graph embedding methods, and it reports the highest accuracy on two hospital datasets.

What carries the argument

The central object is the CBOWRA model: a continuous bag-of-words network whose input layer receives diagnosis vectors and prescription vectors, projects them into embeddings, sums the projections, then performs a dot product with disease vectors before a softmax outputs the probability of each drug being the central word. The ranking accumulation step converts single-sample probabilities into robust anomaly scores by summing central-word probabilities over all combinations of diagnosis context, so that the drug with the lowest accumulated score is the predicted anomaly. The load-bearing assumption is that wrong drugs are less probable as central words in the patient's diagnosis context than correct drugs.

What would settle it

Take a dataset of EHR records with clinician-confirmed medication error labels. Compute CBOWRA's predicted anomaly ranking, then check whether clinically justified off-diagnosis prescriptions, such as laxatives after chemotherapy, are systematically ranked as top anomalies. If a large fraction of clinically appropriate drugs are flagged as anomalies, the core assumption fails.

Watch

Extended reading notes

Core claim

The paper claims that a medication anomaly can be detected as the semantic inconsistency between a patient's diagnoses and prescriptions, and that this reduces to finding the least probable central word under a modified continuous bag-of-words model. Training is expert-free: diagnoses are treated as surrounding context and prescribed drugs as central words, with negative sampling restricted to drugs not prescribed to the patient. At detection time, a ranking accumulation strategy sums each candidate drug's central-word probability across all diagnosis combinations, and the drug with the smallest accumulated rank is the anomaly. On the two hospital datasets the method reaches top-N accuracy of over 0.84 in all settings, with gains of 3.91 to 10.91 percentage points on SG1213 and 0.68 to 2.13 percentage points on XS1213 over the strongest comparison methods.

Load-bearing premise

The method assumes that an anomalous medication is the one least likely to be predicted as the central word given the patient's diagnoses, which breaks for clinically appropriate prescriptions that are unrelated to the listed diagnoses.

Editorial extensions

If this is right

If CBOWRA works as claimed, an expert-free screening tool can rank likely medication errors at the top of a candidate list, letting clinicians focus review effort on fewer records.
The same diagnosis-as-context, drug-as-central-word framing could be applied to other EHR inconsistency checks, such as mismatched lab orders or procedure codes.
Because the model learns semantic consistency from the data itself, it can be retrained on new hospital data without rebuilding feature sets by hand.
The ranking accumulation strategy may transfer to any setting where a model outputs per-candidate scores and the anomaly is the lowest-scoring candidate rather than an outlier in feature space.

Reading between the lines

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

An implicit consequence is that the method's usefulness depends on how often clinically justified off-diagnosis prescriptions occur; the paper's own example of Kaiseru for chemotherapy-induced constipation shows such cases are real, and if they are frequent the model would systematically flag appropriate care as an error.
The method could be tested on a third dataset with known medication error labels from chart review, rather than the pharmacopeia-derived standard answers, to see whether its top-N ranking matches clinician judgments on more nuanced cases.
The dot-product layer between diagnosis vectors and projected prescription embeddings suggests that adding structured clinical knowledge, such as drug indications or side-effect relations, could sharpen the probability contrast without changing the architecture.

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.

Desk Editor's Note

read the letter

The core idea here is not crazy: use CBOW-style embeddings to learn the semantic relation between diagnoses and medications, then rank the least-likely drug as the anomaly. The ranking accumulation post-processing is a simple but sensible trick. The paper also compares against a decent set of baselines (LOF, Naive Bayes variants, TransE/H/D/R) and it openly discusses its data problems in Section IV, which is more than many papers do. But the central claim of superiority does not survive contact with the evaluation.

The biggest problem is that the evaluation is in-sample. No train/validation/test split is described anywhere. The model is trained on the same patient records it then scores, so the reported topN accuracies are training-set fits, not predictive performance. That alone would force me to treat the numbers as upper bounds at best.

The gold standard is worse. It is built by taking each drug's indications from a pharmacopeia website and flagging any drug whose indications do not match the patient's diagnoses. That is exactly the same semantic-mismatch heuristic the method itself learns, so the evaluation risks circularity. More concretely, the paper itself provides counterexamples. The introduction gives the Kaiseru case: a laxative prescribed for chemotherapy-induced constipation, which is correct but would be flagged as an anomaly by the gold standard. Section IV point 1 admits that injections used for hemostasis, analgesia, or side-effect suppression are "classified to be the anomaly, despite the fact they are indeed appropriate medications." Point 2 says glycerol suppositories for chemotherapy constipation are "labeled as an anomaly" by the standard. So the benchmark contains systematic false positives, and the numbers measure agreement with a naive indication-matching rule, not detection of true medication errors.

The reported improvements also do not match the tables. On SG1213, the abstract says gains of 3.91-10.91%; the tables show the CBOWRA advantage over the best baseline ranges from about 5 to 11 percentage points. On XS1213, the abstract claims 0.68-2.13%, but the table shows Top-1 CBOWRA at 0.639 and the best baseline (TransR) at 0.638 — a 0.1 percentage point difference. Calling that a 0.68% improvement is generous; the rest of the columns are similarly tiny. The central claim of consistent superiority is simply not supported by the evidence.

On the positive side, the method is simple, reproducible in principle, and the authors do flag the limitations of their own gold standard in the future work section, acknowledging "low accuracy, incomplete information, and poor timeliness." That honesty earns some credit. But the flaws are load-bearing: the evaluation design invalidates the performance comparison, and the gold standard invalidates the validity of the labels.

Who is this for? Readers working on clinical NLP or EHR anomaly detection might find the problem framing and the CBOW adaptation interesting as a baseline idea, but the empirical claims should not be cited. I would not cite it in my own work, and I would not bring it to a reading group except as an example of how a clever method can be undermined by a weak evaluation.

My recommendation: send it to peer review, because the task is meaningful and the method is salvageable, but the authors should be pushed to use a validated gold standard, a held-out split, and corrected performance numbers. As it stands, I would not accept it.

Referee Report

4 major / 5 minor

Summary. The paper proposes CBOWRA, a representation-learning method to detect medication-diagnosis mismatches in electronic health records. It adapts CBOW so that prescribed drugs are predicted as 'central words' from multi-hot diagnosis contexts; at detection time, each candidate drug receives a probability, scores are ranked ascending, and a ranking-accumulation procedure combines scores over diagnosis subsets, with the lowest accumulated score declared anomalous. The method is evaluated on two Chinese hospital datasets (SG1213, cancer patients; XS1213, heart failure patients) and compared against LOF, Naive Bayes variants, and TransE/H/D/R. The paper reports that CBOWRA achieves the highest TopN accuracy on both datasets, with gains of 3.91-10.91% and 0.68-2.13% over the baselines. The discussion and future-work sections acknowledge limitations of the artificial standard answer set and of modeling only diagnosis and medication concepts.

Significance. If the evaluation were sound, CBOWRA would be a useful, simple, expert-free baseline for flagging candidate medication errors, and the ranking-accumulation idea is a reasonable way to convert softmax probabilities into a detection score. The baseline set is broad, and the paper deserves credit for attempting a comparison across distance-based, probabilistic, and embedding-based methods. However, the central performance claim is not interpretable as stated: the reference standard is the same semantic-mismatch heuristic the model learns, the experimental protocol has no out-of-sample split, and the paper itself supplies examples where the gold standard labels appropriate drugs as anomalies. The reported accuracy margins therefore do not support the claimed superiority, and no code or data is provided to allow independent checking. The novelty of the method itself is modest, but the core idea may merit future work with a valid benchmark.

major comments (4)

[Section III.A.4; Section IV] The reference standard is built by comparing each drug's pharmacopeia indications with the patient's diagnoses and flagging non-matching drugs as abnormal, which encodes the same semantic-mismatch assumption that CBOWRA learns. High agreement with this standard is therefore partly by construction. More seriously, the manuscript itself gives concrete cases where the standard is wrong: Section IV, point 1 states that hemostasis, analgesia, and side-effect-suppression injections are classified as anomalies despite being appropriate, and point 2 states that suppository glycerol is labeled an anomaly although it treats chemotherapy-induced constipation; the introduction's Kaiseru example is the same phenomenon. The reported TopN accuracies thus measure agreement with a naive indication-matching rule rather than detection of true medication errors. The statement that experienced doctors helped ensure accuracy is not supported by any protocol or inter-rater information, and the future-work paragraph concedes that the standard suffers from low accuracy, incomplete information, and poor timeliness.
[Section II.C-D; Section III.A] No train/validation/test split is described anywhere in the experimental setup. The CBOWRA model is trained on the diagnosis-prescription co-occurrences of all patients in each dataset and then used to score those same patients' medications during the ranking-accumulation detection phase. The top-1 values in Table V (0.879) and Table VI (0.639) are therefore in-sample fits, which are likely inflated. An out-of-sample patient-level split, with hyperparameter selection on a validation set and ideally a temporal split, is required before any claim about detection accuracy can be interpreted.
[Section II.B; Section I; Section IV] The detection rule assumes that the probability of a wrong drug appearing in the diagnoses is lower than that of normal drugs and that the medication with the lowest central-word probability is the anomaly. The introduction gives the counterexample of Kaiseru, a laxative appropriate for chemotherapy-induced constipation despite not matching the cancer diagnoses, and Section IV, point 2 reports that the model identifies suppository glycerol as normal while the gold standard labels it anomalous. The paper should quantify how often such clinically appropriate off-diagnosis medications occur in SG1213 and XS1213 and evaluate on clinician-adjudicated labels; otherwise the method will systematically flag correct prescriptions, and its TopN behavior on the current gold standard cannot be interpreted as detection performance.
[Section III.B; Tables V, VI] The sentence claiming that CBOWRA 'possesses the highest accuracy (over 0.84) in all settings' is contradicted by Table VI, where CBOWRA's TopN values range from 0.639 down to 0.622 on XS1213; the 0.84 figure applies only to SG1213. In addition, the paper reports no variance, confidence intervals, or significance tests for any of the comparisons, so the claimed margins over traditional methods are not established.

minor comments (5)

[Table III; Section II.D] The notation is inconsistent: Table III fixes 'No. of Central Words' to 1, while Section II.D instructs selecting k medications as central words and defines s = n - k negative candidates; please clarify k, m, n, and s and how they relate to the Table III parameters.
[Equation (2)] Equation (2) is typeset ambiguously: the division by 2m and the sum over the vocabulary are not clearly structured, so the reader cannot tell whether the denominator is (1/2m) times the sum or the sum of exponentials each scaled by 1/(2m).
[Section II.C] The sentence 'there will be cmk surrounding words' uses undefined symbols c and k; later the text refers to C^m_M without defining whether this is a combination count, and the relationship to the window size m is unclear.
[Section IV] The future-work paragraph says a string similarity matching method was used to construct the standard answers, but Section III.A.4 only describes pharmacopeia indication comparison; please describe the matching procedure or remove the inconsistent reference.
[General] The paper does not state data availability, code availability, or ethics/IRB approval for using real patient EHR data; these details are needed for reproducibility and for assessing compliance in a clinical-data study.

Circularity Check

1 steps flagged · score 6.0 of 10

The reported TopN accuracy is an in-sample fit: the CBOW model is trained and evaluated on the same patient records, with detection performed on 'each training sample,' so the claimed 'predictions' are partly forced by the training data rather than demonstrated out-of-sample.

fitted input called prediction [Section II.C (training-set construction), Section II.D (anomaly detection on training samples), and Section III.A.5 (evaluation with no held-out split)]
"From all the diagnoses of a patient, select m different diagnoses to combine with the medication. In total, there will be cm k surrounding words; 2) Choose one drug from the prescription given to one patient as the central word... For each training sample, we can generate the probability of different medications in the output layer."

The training set is built from every patient's own diagnoses and prescriptions, and the ranking-accumulation step that produces each anomaly score is described as operating on 'each training sample.' No train/test split is reported in Section III, so the TopN accuracies in Tables V and VI measure the model's reconstruction of the same diagnosis-medication pairs it was optimized on. The probabilities that decide whether a drug is the 'least possible central word' come from embeddings trained on those very patients, so each evaluated patient's ranking is an in-sample output rather than a prediction. The reported superiority (e.g., Top-1 0.879 on SG1213, 0.639 on XS1213) is therefore partly forced by construction and does not establish generalization to unseen records.

full rationale

The only concrete reduction I can exhibit from the paper's own text is the absence of any train/test separation. CBOWRA is an unsupervised representation learner and is not fitted to the anomaly labels, so the gold-standard construction is not a fitted-parameter circularity. The gold standard is independently problematic on correctness grounds: Section IV itself states that appropriate medications such as glycerol suppositories 'are labeled as an anomaly,' and the paper admits the standard answer construction has 'low accuracy, incomplete information, and poor timeliness of medical knowledge.' That is a benchmark-validity concern rather than a derivation-chain circularity. There is no load-bearing self-citation: the word2vec citations [24] and [29] are standard external sources, and the method's design is not justified by a same-author uniqueness theorem. The central circularity is that detection is performed on the same 'training samples' from which the CBOW distribution was learned, so the reported TopN numbers are in-sample fits. Had a held-out split been used, the in-sample issue would disappear, leaving only the separate benchmark-validity problem; as written, the main empirical claim is evaluated on the training data, making the 'prediction' partly a fitted score.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new entities. The central claim relies on two domain assumptions about rarity and learnability of anomalies, plus a noisy gold standard, and on several under-specified hyperparameters that were tuned on the same data used for evaluation.

free parameters (5)

Number of surrounding words (window size m) = 2
Set after evaluating values 1 to 10 on the same data; the paper says performance was stable after k greater than 3 but still chose 2.
Number of negative samples = 5
Chosen without a stated justification or sensitivity analysis.
Number of central words = 1
Set in the parameter table, but the algorithm description in Section II.D refers to k central words, which is inconsistent.
Number of diseases = 64
Parameter table lists No. of Diseases 64 but the two datasets have 440 and 643 diagnoses; this value is unexplained and likely refers to a processed subset.
Embedding dimension = not specified
The paper mentions a 300 by 20000 matrix in a general word2vec discussion, but the actual embedding size used in experiments is not reported.

assumptions (3)

domain assumption Anomalous medications are rare and have lower conditional probability than correct medications given the patient's diagnoses.
Adopted in Section II.B: 'it is reasonable to assume that the probability of wrong drug appearing in the diagnoses is lower than the normal or correct drugs.' This is load-bearing because the detection rule flags the lowest-probability drug.
domain assumption The semantic relation between diagnoses and prescriptions is learnable by a CBOW-style model with a dot-product modification.
The whole method assumes that a shallow neural model trained on co-occurrence can capture the semantic consistency between diagnoses and drugs; this is standard but unproven for this clinical corpus.
domain assumption The standard answer set built from pharmacopeia indications and string matching is accurate enough to evaluate detection.
The gold standard is used for all evaluation; the authors themselves note it may have low accuracy and incomplete information in Section IV.

how reviews work

0 comments

Cite this review

Pith. "Pith review of CBOWRA: A Representation Learning Approach for Medication Anomaly Detection." pith.science (2026). https://pith.science/paper/KIAYSIMV

@misc{pith2026190807147,
  author       = {Pith},
  title        = {Pith review of: CBOWRA: A Representation Learning Approach for Medication Anomaly Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KIAYSIMV}},
  note         = {Machine review of arXiv:1908.07147}
}

read the original abstract

Electronic health record is an important source for clinical researches and applications, and errors inevitably occur in the data, which could lead to severe damages to both patients and hospital services. One of such error is the mismatches between diagnoses and prescriptions, which we address as 'medication anomaly' in the paper, and clinicians used to manually identify and correct them. With the development of machine learning techniques, researchers are able to train specific model for the task, but the process still requires expert knowledge to construct proper features, and few semantic relations are considered. In this paper, we propose a simple, yet effective detection method that tackles the problem by detecting the semantic inconsistency between diagnoses and prescriptions. Unlike traditional outlier or anomaly detection, the scheme uses continuous bag of words to construct the semantic connection between specific central words and their surrounding context. The detection of medication anomaly is transformed into identifying the least possible central word based on given context. To help distinguish the anomaly from normal context, we also incorporate a ranking accumulation strategy. The experiments were conducted on two real hospital electronic medical records, and the topN accuracy of the proposed method increased by 3.91 to 10.91% and 0.68 to 2.13% on the datasets, respectively, which is highly competitive to other traditional machine learning-based approaches.

Figures

Figures reproduced from arXiv: 1908.07147 by the authors.

**Figure 4.** The network structure of CBOW. The inputs Wt−n, Wt−n+1, · · · , Wt+n denote the surrounding words that are presented in vector forms. The projection layer does a vector-wise summation of all inputs and passes the result to a softmax layer to produce the probability of each candidate word being the central word. The training of the model can be formulated as maximizing the probability of generating a correct central… view at source ↗

**Figure 3.** Examples of vectors for patients’ prescriptions. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 5.** Figure 5: Network structure of CBOWRA. denote the surrounding words and MT ar, · · · , Mneg are the prescriptions given to the patient. Only surrounding words are transformed into word embeddings, before passing through a projection layer. The projection layer works in the same …

**Figure 6.** Figure 6: Examples of ranking scores. To tackle the problem, we applied a ranking strategy, called Ranking Accumulation (RA), in the process of anomaly detection. For a specific diagnosis, it is intuitive that the smaller the probability that a drug is predicted to be a normal …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 25 canonical work pages

[1]

Mining electronic health records (EHRs): a survey,

P. Yadav, M. Steinbach, V . Kumar, and G. Simon, “Mining electronic health records (EHRs): a survey,” ACM Computing Surveys (CSUR) , vol. 50, no. 6, p. 85, 2018

work page 2018
[2]

PatientEG Dataset: Bringing Event Graph Model with Temporal Relations to Electronic Medical Records

X. Liu, J. Jin, Q. Wang, T. Ruan, Y . Zhou, D. Gao, and Y . Yin, “Patienteg dataset: Bringing event graph model with temporal relations to electronic medical records,” arXiv preprint arXiv:1812.09905 , 2018

work page Pith review arXiv 2018
[3]

The challenges in deﬁning and measuring diagnostic error,

L. Zwaan and H. Singh, “The challenges in deﬁning and measuring diagnostic error,” Diagnosis, vol. 2, no. 2, pp. 97–103, 2015

work page 2015
[4]

Medication errors,

C. Galper, “Medication errors,” The Lancet , vol. 349, no. 9056, pp. 959–960, 1997

work page 1997
[5]

Diagnostic error in medicine: analysis of 583 physician-reported errors,

G. D. Schiff, O. Hasan, S. Kim, R. Abrams, K. Cosby, B. L. Lambert, A. S. Elstein, S. Hasler, M. L. Kabongo, N. Krosnjar et al., “Diagnostic error in medicine: analysis of 583 physician-reported errors,” Archives of internal medicine , vol. 169, no. 20, pp. 1881–1887, 2009

work page 2009
[6]

Outlier analysis,

C. C. Aggarwal, “Outlier analysis,” in Data mining . Springer, 2015, pp. 237–263

work page 2015
[7]

Distance-based outlier detection in data streams,

L. Tran, L. Fan, and C. Shahabi, “Distance-based outlier detection in data streams,” Proceedings of the VLDB Endowment , vol. 9, no. 12, pp. 1089–1100, 2016

work page 2016
[8]

Privacy-preserving lof outlier detection,

L. Li, L. Huang, W. Yang, X. Yao, and A. Liu, “Privacy-preserving lof outlier detection,” Knowledge and Information Systems , vol. 42, no. 3, pp. 579–597, 2015

work page 2015

Show all 29 references

[9]

On the evaluation of outlier detection and one-class classiﬁcation methods,

L. Swersky, H. O. Marques, J. Sander, R. J. Campello, and A. Zimek, “On the evaluation of outlier detection and one-class classiﬁcation methods,” in IEEE International Conference on Data Science and Advanced Analytics (DSAA) . IEEE, 2016, pp. 1–10

2016
[10]

On the evaluation of unsu- pervised outlier detection: measures, datasets, and an empirical study,

G. O. Campos, A. Zimek, J. Sander, R. J. Campello, B. Micenkov ´a, E. Schubert, I. Assent, and M. E. Houle, “On the evaluation of unsu- pervised outlier detection: measures, datasets, and an empirical study,” Data Mining and Knowledge Discovery , vol. 30, no. 4, pp. 891–927, 2016

2016
[11]

A com- parative evaluation of outlier detection algorithms: Experiments and analyses,

R. Domingues, M. Filippone, P. Michiardi, and J. Zouaoui, “A com- parative evaluation of outlier detection algorithms: Experiments and analyses,” Pattern Recognition, vol. 74, pp. 406–421, 2017

2017
[12]

Outlier detection in healthcare fraud: A case study in the medicaid dental domain,

G. V . Capelleveen, M. Poel, R. M. Mueller, D. Thornton, and J. V . Hillegersberg, “Outlier detection in healthcare fraud: A case study in the medicaid dental domain,” International Journal of Accounting Information Systems, vol. 21, pp. 18–31, 2016

2016
[13]

Medication errors: new eu good practice guide on risk minimisation and error prevention,

T. Goedecke, K. Ord, V . Newbould, S. Brosch, and P. Arlett, “Medication errors: new eu good practice guide on risk minimisation and error prevention,” Drug safety, vol. 39, no. 6, pp. 491–500, 2016

2016
[14]

Screening for medication errors using an outlier detection system,

G. D. Schiff, L. A. V olk, M. V olodarskaya, D. H. Williams, L. Walsh, S. G. Myers, D. W. Bates, and R. Rozenblum, “Screening for medication errors using an outlier detection system,” Journal of the American Medical Informatics Association , vol. 24, no. 2, pp. 281–287, 2017

2017
[15]

Probabilistic-mismatch anomaly detection: Do one’s med- ications match with the diagnoses,

L. Zhang, X. Li, H. Liu, J. Mei, G. Hu, J. Zhao, Y . Zou, B. Xie, and G. Xie, “Probabilistic-mismatch anomaly detection: Do one’s med- ications match with the diagnoses,” in 2016 IEEE 16th International Conference on Data Mining (ICDM) . IEEE, 2016, pp. 659–668

2016
[16]

Predicting urinary tract infections in the emergency department with machine learning,

R. A. Taylor, C. L. Moore, K.-H. Cheung, and C. Brandt, “Predicting urinary tract infections in the emergency department with machine learning,” PloS one, vol. 13, no. 3, p. e0194085, 2018

2018
[17]

Drug prescription support in dental clinics through drug corpus mining,

W. P. Goh, X. Tao, J. Zhang, J. Yong, W. Zhang, and H. Xie, “Drug prescription support in dental clinics through drug corpus mining,” International Journal of Data Science and Analytics , vol. 6, no. 4, pp. 341–349, 2018

2018
[18]

Natural language processing tech- nologies in radiology research and clinical applications,

T. Cai, A. A. Giannopoulos, S. Yu, T. Kelil, B. Ripley, K. K. Kumamaru, F. J. Rybicki, and D. Mitsouras, “Natural language processing tech- nologies in radiology research and clinical applications,” Radiographics, vol. 36, no. 1, pp. 176–191, 2016

2016
[19]

Natural language processing in radiology: a systematic review,

E. Pons, L. M. Braun, M. M. Hunink, and J. A. Kors, “Natural language processing in radiology: a systematic review,”Radiology, vol. 279, no. 2, pp. 329–343, 2016

2016
[20]

Learning semantic textual similarity from conversations,

Y . Yang, S. Yuan, D. Cer, S.-y. Kong, N. Constant, P. Pilar, H. Ge, Y .-H. Sung, B. Strope, and R. Kurzweil, “Learning semantic textual similarity from conversations,” arXiv preprint arXiv:1804.07754 , 2018

2018 arXiv
[21]

Video captioning with attention-based lstm and semantic consistency,

L. Gao, Z. Guo, H. Zhang, X. Xu, and H. T. Shen, “Video captioning with attention-based lstm and semantic consistency,” IEEE Transactions on Multimedia, vol. 19, no. 9, pp. 2045–2055, 2017

2017
[22]

Learning semantic representa- tions for unsupervised domain adaptation,

S. Xie, Z. Zheng, L. Chen, and C. Chen, “Learning semantic representa- tions for unsupervised domain adaptation,” in International Conference on Machine Learning , 2018, pp. 5419–5428

2018
[23]

Fast and accurate recognition of chinese clinical named entities with residual dilated convolutions,

J. Qiu, Q. Wang, Y . Zhou, T. Ruan, and J. Gao, “Fast and accurate recognition of chinese clinical named entities with residual dilated convolutions,” in2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) . IEEE, 2018, pp. 935–942

2018
[24]

Efﬁcient estimation of word representations in vector space,

T. Mikolov, K. Chen, G. Corrado, and J. Dean, “Efﬁcient estimation of word representations in vector space,” arXiv preprint arXiv:1301.3781 , 2013

2013 arXiv
[25]

Translating embeddings for modeling multi-relational data,

A. Bordes, N. Usunier, A. Garcia-Duran, J. Weston, and O. Yakhnenko, “Translating embeddings for modeling multi-relational data,” in Ad- vances in neural information processing systems , 2013, pp. 2787–2795

2013
[26]

Knowledge graph embedding by translating on hyperplanes,

Z. Wang, J. Zhang, J. Feng, and Z. Chen, “Knowledge graph embedding by translating on hyperplanes,” in Twenty-Eighth AAAI conference on artiﬁcial intelligence, 2014

2014
[27]

Knowledge graph embedding via dynamic mapping matrix,

G. Ji, S. He, L. Xu, K. Liu, and J. Zhao, “Knowledge graph embedding via dynamic mapping matrix,” in Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Lon...

2015
[28]

Learning entity and relation embeddings for knowledge graph completion,

Y . Lin, Z. Liu, M. Sun, Y . Liu, and X. Zhu, “Learning entity and relation embeddings for knowledge graph completion,” in Twenty-ninth AAAI conference on artiﬁcial intelligence , 2015

2015
[29]

Distributed representations of words and phrases and their composi- tionality,

T. Mikolov, I. Sutskever, K. Chen, G. S. Corrado, and J. Dean, “Distributed representations of words and phrases and their composi- tionality,” in Advances in neural information processing systems , 2013, pp. 3111–3119

2013

Pith tools

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

[1] [1]

Mining electronic health records (EHRs): a survey,

P. Yadav, M. Steinbach, V . Kumar, and G. Simon, “Mining electronic health records (EHRs): a survey,” ACM Computing Surveys (CSUR) , vol. 50, no. 6, p. 85, 2018

work page 2018

[2] [2]

PatientEG Dataset: Bringing Event Graph Model with Temporal Relations to Electronic Medical Records

X. Liu, J. Jin, Q. Wang, T. Ruan, Y . Zhou, D. Gao, and Y . Yin, “Patienteg dataset: Bringing event graph model with temporal relations to electronic medical records,” arXiv preprint arXiv:1812.09905 , 2018

work page Pith review arXiv 2018

[3] [3]

The challenges in deﬁning and measuring diagnostic error,

L. Zwaan and H. Singh, “The challenges in deﬁning and measuring diagnostic error,” Diagnosis, vol. 2, no. 2, pp. 97–103, 2015

work page 2015

[4] [4]

Medication errors,

C. Galper, “Medication errors,” The Lancet , vol. 349, no. 9056, pp. 959–960, 1997

work page 1997

[5] [5]

Diagnostic error in medicine: analysis of 583 physician-reported errors,

G. D. Schiff, O. Hasan, S. Kim, R. Abrams, K. Cosby, B. L. Lambert, A. S. Elstein, S. Hasler, M. L. Kabongo, N. Krosnjar et al., “Diagnostic error in medicine: analysis of 583 physician-reported errors,” Archives of internal medicine , vol. 169, no. 20, pp. 1881–1887, 2009

work page 2009

[6] [6]

Outlier analysis,

C. C. Aggarwal, “Outlier analysis,” in Data mining . Springer, 2015, pp. 237–263

work page 2015

[7] [7]

Distance-based outlier detection in data streams,

L. Tran, L. Fan, and C. Shahabi, “Distance-based outlier detection in data streams,” Proceedings of the VLDB Endowment , vol. 9, no. 12, pp. 1089–1100, 2016

work page 2016

[8] [8]

Privacy-preserving lof outlier detection,

L. Li, L. Huang, W. Yang, X. Yao, and A. Liu, “Privacy-preserving lof outlier detection,” Knowledge and Information Systems , vol. 42, no. 3, pp. 579–597, 2015

work page 2015