Pith. sign in

REVIEW 5 major objections 7 minor 34 references

CDE-Mapper: Using Retrieval-Augmented Language Models for Linking Clinical Data Elements to Controlled Vocabularies

T0 review · 5 major / 7 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper introduces CDE-Mapper, a retrieval-augmented language-model pipeline that links clinical data elements to controlled vocabularies and reports top-1 accuracy above dedicated biomedical matchers on four datasets.

desk verdict A plausible RAG system for clinical data element linking, but the unreported state of the knowledge reservoir during evaluation could turn its headline accuracies into lookup rather than linking. read the letter →

arxiv 2505.04365 v1 pith:3MEWUNRY submitted 2025-05-07 cs.IR

Komal Gilani , Marlo Verket , Christof Peters , Michel Dumontier , Hans-Peter Brunner-La Rocca , Visara Urovi This is my paper · ORCID
classification cs.IR
keywords clinicaldataelementsconceptlinkingretrieval-augmentedgenerationlargelanguagemodelscontrolledvocabulariesentitynormalizationOMOPcommonmodelharmonization
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

Clinical data elements—variables in data dictionaries such as "heart rate measured in recumbent position"—are recorded differently across systems, and standardizing them by linking each element to a controlled vocabulary (SNOMED, LOINC, RxNorm, and similar) is a bottleneck for interoperability. The paper claims that a modular retrieval-augmented LLM pipeline, called CDE-Mapper, solves this for both simple atomic elements and complex composite ones. The paper reports that CDE-Mapper with Llama3.1 or GPT4o-mini beats established biomedical matchers on four datasets, reaching 94.4% top-1 accuracy on NCBI-DC and 86.4% on heart-failure data, an average of 7.2 percentage points higher than baselines. The design matters because composite elements, which encode several attributes at once, are exactly where earlier matchers fail.

What carries the argument

The load-bearing mechanism is the modular RAG pipeline. It converts each input row into a JSON query decomposition with the LLM, retrieves candidates from a three-million-concept knowledge base using SPLADE (sparse, for exact canonical names) and SapBERT (dense, for semantic similarity), discards candidates below a cosine-similarity threshold, then re-ranks the survivors with an LLM that scores each candidate from 1 to 10 and classifies it, repeated three times with a binary confidence threshold $\tau = 0.85n$; only consistently highly-ranked candidates win. The knowledge reservoir, a dictionary or triple store of expert-validated label-concept pairs with OMOP IDs, lets later occurrences of the same label skip the whole retrieval-and-reranking path.

What would settle it

Clear the knowledge reservoir before evaluating on NCBI-DC and HF Studies and draw in-context examples only from outside those test sets; if top-1 accuracy falls well below the reported 94.4% and 86.4%, part of the reported gain comes from cached or exemplified test mappings rather than from the linking pipeline itself.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that decomposing a clinical data element into structured subqueries (base entity, associated entities, categories, unit, visit), retrieving candidate vocabulary concepts with both dense and sparse retrievers, filtering those candidates by similarity, and re-ranking the survivors with an LLM yields higher concept-linking accuracy than fine-tuned biomedical encoders or a prompt-only LLM baseline. The strongest results come from the Llama3.1 variant, with 94.4% accuracy on NCBI-DC and 86.4% on the composite-heavy HF Studies dataset. The framework also introduces a knowledge reservoir: correctly mapped label-concept pairs, validated by an LLM judge and then by clinicians, are stored for direct reuse, cutting inference cost on later queries.

Load-bearing premise

The evaluation assumes the knowledge reservoir and in-context examples contain no test-set gold-standard mappings—in particular, that each dataset run starts with an empty reservoir, so the reported top-1 accuracy reflects the model's linking ability rather than retrieval of expert-validated answers.

Editorial extensions

If this is right

  • Clinical data entries that bundle several attributes, such as a family-history field or a biomarker with a time point and measurement method, can be standardized without being split into separate records by hand.
  • Because the reservoir reuses validated mappings, the system's per-query inference cost should fall as it is used, making large-dictionary standardization feasible in practice.
  • Open-source Llama3.1 matches or beats the proprietary GPT-4 variant in most settings, so the method does not depend on costly closed models.
  • Ablations attribute measurable gains to context-aware retrieval with knowledge filtering and to the second reranking step, implying that retriever quality and reranking matter more than the choice of LLM alone.

Reading between the lines

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

  • If the evaluation reservoir is empty at the start of each dataset run, then the reported accuracy is a cold-start number and would likely rise as the reservoir accumulates validated mappings; a warm-start deployment should be at least as accurate.
  • The same JSON decomposition could be applied to free-text clinical notes, not only structured dictionaries, by letting the LLM first extract attribute-value pairs from a sentence and then linking each pair.
  • Because gains were smaller on BC5CDR-D, where few concepts have synonyms or parent terms, the pipeline's accuracy should be predictable from vocabulary richness; enriching sparse vocabularies is a cheap way to test that prediction.
  • A long-tail analysis would be a natural stress test: reranking with a high confidence threshold may favor frequent concepts, so accuracy on rare diseases could differ from overall accuracy.
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

5 major / 7 minor

Summary. The paper proposes CDE-Mapper, a modular retrieval-augmented generation framework for linking clinical data elements (CDEs) to controlled vocabularies such as SNOMED, LOINC, and RxNorm. The pipeline consists of query decomposition into structured subqueries, ensemble retrieval with SapBERT and SPLADE embeddings, knowledge filtering, a two-step LLM reranking module, and a knowledge reservoir intended to cache validated mappings for future queries. The authors evaluate CDE-Mapper on four datasets (BC5CDR-Disease, NCBI-DC, MIID, and HF Studies) and report top-1 accuracy improvements over SapBERT, KRISS BERT, BioBERT-snomed, and PromptLink, with the highest accuracy of 94.4% on NCBI-DC and 86.4% on HF Studies. Ablation studies examine query decomposition, knowledge filtering, reranking, and performance by CDE type, and a case study compares predictions with clinician judgments.

Significance. If the reported results are valid, the paper makes a useful practical contribution: it addresses composite and dependent CDEs, which are often neglected in entity-linking benchmarks, and it provides a modular architecture that could be adapted to other clinical harmonization tasks. The paper also includes ablation studies and a clinician-in-the-loop case study, which are strengths. However, the headline accuracy claims are currently difficult to interpret because the evaluation protocol for the knowledge reservoir is underspecified; depending on how the reservoir was initialized and updated, the reported numbers could reflect lookup from previously validated mappings rather than the retrieval-and-reranking pipeline. The statistical support for the significance claims is also incomplete. The framework itself is plausible, and the issues are addressable with additional experimental detail and, if necessary, reruns with a cold reservoir.

major comments (5)
  1. [§2.7.6, Algorithm 1 (Appendix A)] The state and role of the knowledge reservoir during evaluation is never specified. Algorithm 1 checks CheckInKR(qi) before retrieval (line 8) and calls AddInKR(qi, candidate_result[qi]) for every processed component (line 26), whereas §2.7.6 states that only expert-validated concepts are stored. The paper does not state whether the reservoir was empty at the start of each dataset run, whether it was pre-populated with validated gold-standard mappings, or how the human-in-the-loop validation was applied during the reported experiments. If the reservoir was warm with test or gold mappings, the reported acc@1 values in Table 3 could reflect exact-label lookup rather than the retrieval-reranking pipeline; even a cold reservoir that accumulates every prediction can answer repeated mentions from earlier model outputs. This ambiguity is especially consequential for NCBI-DC, which has 73,024 mentions but only 359 unique concepts (Table 1). Please specify the exact evaluation protocol: reservoir initialization, the validation step applied during evaluation, and whether results are reported with the reservoir enabled; ideally report both cold and warm configurations.
  2. [Table 3, §4] The claim of statistical significance (†, T-test, p<0.05) is not supported by the reported information: no confidence intervals, standard deviations, number of runs, or unit of analysis (mentions, unique concepts, or datasets) are given. With deterministic retrieval and LLM inference, it is also unclear what variability the test captures. Please provide the full statistical details or remove the significance claim; note also that BC5CDR-D entries are not marked † even though the text describes improvements there.
  3. [§2.7.7, Table 7] The reranking threshold τ=0.85×n is chosen based on "preliminary work (results not shown)", and no sensitivity analysis is reported. Since this threshold directly determines which candidates are accepted and thereby the acc@1 numbers in Table 7, the paper should report the tuning procedure, the range of thresholds tried, and the effect of τ on the results; otherwise the reported gains from reranking cannot be reproduced or attributed.
  4. [§2.7.6, §5] The knowledge reservoir is presented as a contribution that "minimizes computational costs", but no experiment measures its effect on accuracy or inference time/latency. The ablation studies in §4.2 cover retrieval, filtering, and reranking, but not the reservoir. Please add an evaluation that isolates the reservoir's contribution, such as cache hit rate, latency reduction, and accuracy with the reservoir disabled versus enabled.
  5. [Table 4, §4.2.1] The per-type CDE results (atomic, composite, dependent) are reported without sample sizes or confidence intervals; since only the HF dataset (476 unique concepts) contains composite CDEs, the comparison for composite and dependent CDEs rests on a small subset. Please report the number of elements per category and appropriate variability measures so that the composite-CDE claims can be evaluated.
minor comments (7)
  1. [Throughout] There are numerous typos and formatting errors: "sectionMethods" appears before §2.6, the appendix heading reads "A Alogrithm", §3 has "consine similarity alogirthm", §4.3 has "clinicasn", and the Figure 1 caption contains "for for". The term "infraction" is used where "infarction" is intended in several places. A careful proofreading pass is needed.
  2. [Table 4] The column header "CDE-Mapper SapBERT KRISS BERT BioBERT-snomedLlama3.1 GPT4o-mini GPT4" is garbled, making the alignment of the numeric values ambiguous. Please reformat the table so that each model variant has a distinct, clearly labeled column.
  3. [§5, Table 1] The discussion states "Of 73,412 concepts, only 6,530 had hierarchical information" for a dataset, but Table 1 lists BC5CDR-D as having 73,126 mentions. Please clarify whether the number refers to mentions, concepts, or a different collection, and reconcile the two figures.
  4. [Algorithm 1] The indentation of AddInKR (line 26) is ambiguous: as printed, it could be read as executing even when candidates is empty, storing "NA" as a reservoir entry. Please make the control flow explicit so that it is clear under which branches the reservoir is updated.
  5. [References] Some references are incomplete: the MetaMapLite entry is just "Lhncbc" with a URL, and the OHDSI Ananke citation lacks full author and venue details. Please verify all bibliographic entries for completeness.
  6. [Table 8, §4.3] The case study would benefit from a description of how the clinician judgments were collected, including the number of clinicians, whether they were blinded to model predictions, and how disagreements were resolved, to support the symbol denoting "predictions considered correct by clinicians".
  7. [General] The paper does not include a data or code availability statement. Given the reproducibility concerns raised above, please add one or explain why the artifacts cannot be shared.

Circularity Check

1 steps flagged · score 5.0 of 10

Knowledge reservoir can convert repeated test labels into cached lookups; NCBI-DC has 73,024 mentions but 359 concepts, so a large share of the reported acc@1 may be reservoir hits rather than RAG linking.

  1. fitted input called prediction [Algorithm 1, lines 8 and 26; Section 2.7.6 ('Knowledge Reservoir'); Section 3 ('Experiment Design')]
    "kr_match← CheckInKR(qi) ▷ Check if candidates exists in Knowledge Reservoir ... AddInKR(qi, candidate_result[qi]) ▷ Add new candidate to Knowledge Reservoir ... The reservoir is dynamically updated as new knowledge is retrieved ... The datasets presented in Table. 1 were used as test entities / variables for concept linking."

    The benchmark datasets are used directly as test queries, and Algorithm 1 checks the reservoir before any retrieval (line 8), returning the stored candidate as the result (line 10: candidate_result[qi]←kr_match). Line 26 then unconditionally stores the produced candidate during the same evaluation pass. Therefore, once a label has been processed, every later mention of that label is answered by the stored value, not by the RAG pipeline; the output equals the previously stored answer by construction. Table 1 reports NCBI-DC with 73,024 mentions but only 359 unique concepts, so a large fraction of the 94.4% acc@1 can be cache hits on repeated labels. The paper never states whether the reservoir was empty at the start of each dataset run or pre-populated with expert-validated test mappings.

full rationale

The modular CDE-Mapper pipeline itself is not circular: query decomposition, ensemble retrieval (SapBERT, SPLADE), knowledge filtering, and two-step LLM reranking are described independently, and the ablation tables provide internal evidence that each component contributes. The comparison baselines (SapBERT, KRISS BERT, BioBERT-snomed, PromptLink) are external published methods, and no load-bearing result is justified solely by a self-citation chain. The threshold tau=0.85n is said to be chosen from 'preliminary work (results not shown)', which is a hyperparameter-tuning concern rather than a derivation-level circularity. The significant circularity risk is evaluation-level: Algorithm 1's unconditional AddInKR during testing, combined with CheckInKR before retrieval, means the reservoir is both the store of prior answers and the source of future answers within the same test set. Given the heavily skewed mention-to-concept ratio in NCBI-DC, the 94.4% acc@1 figure cannot be interpreted as pure linking accuracy without knowing whether the reservoir was cold, warm, or validated on test labels. This is a partial circularity in the reported empirical claim, not in the methodological derivation, hence the score of 5 rather than a higher value.

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

The central claim rests on domain assumptions about vocabulary synonym coverage, expert linking rules, and the independence of human validation, plus several tuned thresholds. No new physical or formal entities are introduced; the knowledge reservoir is a software component that stores validated label-concept pairs, and its evaluation-time state is the main unresolved circularity risk.

free parameters (4)
  • knowledge filter similarity threshold tau
    Defined in Section 2.7.5 as adjustable; no value or sensitivity analysis is reported.
  • reranking confidence threshold tau = 0.85 * n
    Section 2.7.7 sets tau = 0.85 * n based on 'preliminary work (results not shown)'.
  • number of retrieved candidates k = 10
    Section 3 selects k=10 citing common practice; no ablation on k is reported.
  • number of self-consistency prompts n = 3
    Section 2.7.7 uses n=3 prompting rounds; the choice is not justified by sensitivity analysis.
assumptions (4)
  • domain assumption Expert-defined linking rules map conditions to SNOMED, drugs to RxNorm or ATC, labs to LOINC, and historical context to past-condition concepts.
    Section 2.7.2 treats these rules as correct and complete; systematic errors here propagate to all links.
  • domain assumption Synonym expansion via cross-vocabulary equivalence relationships improves rather than distorts retrieval.
    Section 2.7.1 merges synonyms from mapped vocabularies; no evaluation of precision loss from noisy synonym expansion is given.
  • domain assumption Human expert validation of the knowledge reservoir is independent of the model outputs and represents ground truth.
    Section 2.7.6 relies on experts to validate LLM-judged concepts; independence from test labels is not documented.
  • domain assumption Dense and sparse embedding spaces (SapBERT, SPLADE) contain the correct concept within the top retrieved candidates.
    Retrieval and reranking can only select among returned candidates; no recall analysis of the retriever against the full knowledge base is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CDE-Mapper: Using Retrieval-Augmented Language Models for Linking Clinical Data Elements to Controlled Vocabularies." pith.science (2026). https://pith.science/paper/3MEWUNRY

@misc{pith2026250504365,
  author       = {Pith},
  title        = {Pith review of: CDE-Mapper: Using Retrieval-Augmented Language Models for Linking Clinical Data Elements to Controlled Vocabularies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3MEWUNRY}},
  note         = {Machine review of arXiv:2505.04365}
}
read the original abstract

The standardization of clinical data elements (CDEs) aims to ensure consistent and comprehensive patient information across various healthcare systems. Existing methods often falter when standardizing CDEs of varying representation and complex structure, impeding data integration and interoperability in clinical research. We introduce CDE-Mapper, an innovative framework that leverages Retrieval-Augmented Generation approach combined with Large Language Models to automate the linking of CDEs to controlled vocabularies. Our modular approach features query decomposition to manage varying levels of CDEs complexity, integrates expert-defined rules within prompt engineering, and employs in-context learning alongside multiple retriever components to resolve terminological ambiguities. In addition, we propose a knowledge reservoir validated by a human-in-loop approach, achieving accurate concept linking for future applications while minimizing computational costs. For four diverse datasets, CDE-Mapper achieved an average of 7.2\% higher accuracy improvement compared to baseline methods. This work highlights the potential of advanced language models in improving data harmonization and significantly advancing capabilities in clinical decision support systems and research.

Figures

Figures reproduced from arXiv: 2505.04365 by the authors.

Figure 1
Figure 1. Atomic and Composite terms for for 4 example clinical data elements (sex, GGT, FHH, Caisson). The CDEs [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Modular Framework Stepwise Workflow for Concept Linking of Common Data Elements Using RAG [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Detailed Overview of Proposed Modular RAG Framework for normalization and linking CDEs in controlled [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Two-step reranking to select a candidate using a combination of classification and relevance score. [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Comparison of NCGD@K for Different Models Across datasets [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 13 canonical work pages

  1. [5]

    SNOBERT: A Benchmark for clinical notes entity linking in the SNOMED CT clinical terminology

    Mikhail Kulyabin, Gleb Sokolov, Aleksandr Galaida, Andreas Maier, and Tomas Arias-Vergara. Snobert: A benchmark for clinical notes entity linking in the snomed ct clinical terminology. arXiv preprint arXiv:2405.16115,

    work page Pith review arXiv

  2. [6]

    Daniel Loureiro and A

    doi:10.1109/ACCESS.2021.3119621. Daniel Loureiro and A. Jorge. Medlinker: Medical entity linking with neural representations and dictionary matching. Advances in Information Retrieval, 12036:230 – 237,

    work page arXiv 2021

  3. [7]

    Nadeesha Perera, M

    doi:10.1007/978-3-030-45442-5_29. Nadeesha Perera, M. Dehmer, and F. Emmert-Streib. Named entity recognition and relation detection for biomedical information extraction. Frontiers in Cell and Developmental Biology, 8,

    work page doi:10.1007/978-3-030-45442-5_29

  4. [8]

    Tassallah Abdullahi, Laura Mercurio, Ritambhara Singh, and Carsten Eickhoff

    doi:10.3389/fcell.2020.00673. Tassallah Abdullahi, Laura Mercurio, Ritambhara Singh, and Carsten Eickhoff. Retrieval-based diagnostic decision support: Mixed methods study. JMIR Medical Informatics, 12:e50209,

    work page arXiv 2020

  5. [10]

    A comprehensive overview of large language models

    Humza Naveed, Asad Ullah Khan, Shi Qiu, Muhammad Saqib, Saeed Anwar, Muhammad Usman, Naveed Akhtar, Nick Barnes, and Ajmal Mian. A comprehensive overview of large language models. arXiv preprint arXiv:2307.06435,

    arXiv

  6. [11]

    Large language models in medicine

    Arun James Thirunavukarasu, Darren Shu Jeng Ting, Kabilan Elangovan, Laura Gutierrez, Ting Fang Tan, and Daniel Shu Wei Ting. Large language models in medicine. Nature medicine, 29(8):1930–1940,

    work page 1930

  7. [13]

    doi:10.1093/bioinformatics/btae104

    ISSN 1367-4811. doi:10.1093/bioinformatics/btae104. URL http://dx.doi.org/10.1093/bioinformatics/btae104. Yue Zhang, Yafu Li, Leyang Cui, Deng Cai, Lemao Liu, Tingchen Fu, Xinting Huang, Enbo Zhao, Yu Zhang, Yulong Chen, Longyue Wang, A. Luu, Wei Bi, Freda Shi, and Shuming Shi. Siren’s song in the ai ocean: A survey on hallucination in large language mode...

    arXiv

  8. [14]

    Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, Meng Wang, and Haofen Wang

    doi:10.48550/arXiv.2309.01219. Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, Meng Wang, and Haofen Wang. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997,

    arXiv doi:10.48550/arxiv.2309.01219

Show all 34 references
  1. [15]

    Fangyu Liu, Ehsan Shareghi, Zaiqiao Meng, Marco Basaldella, and Nigel Collier

    URL https://arxiv.org/abs/2312.10997. Fangyu Liu, Ehsan Shareghi, Zaiqiao Meng, Marco Basaldella, and Nigel Collier. Self-alignment pretraining for biomedical entity representations. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Com...

    2021 arXiv

  2. [16]

    Knowledge-rich self-supervision for biomedical entity linking

    Sheng Zhang, Hao Cheng, Shikhar Vashishth, Cliff Wong, Jinfeng Xiao, Xiaodong Liu, Tristan Naumann, Jianfeng Gao, and Hoifung Poon. Knowledge-rich self-supervision for biomedical entity linking. arXiv preprint arXiv:2112.07887,

    arXiv

  3. [17]

    The fda sentinel initiative—an evolving national resource

    Richard Platt, Jeffrey S Brown, Melissa Robb, Mark McClellan, Robert Ball, Michael D Nguyen, and Rachel E Sherman. The fda sentinel initiative—an evolving national resource. N Engl J Med, 379(22):2091–2093,

  4. [18]

    Few-shot learning with retrieval augmented language models

    Gautier Izacard, Patrick Lewis, Maria Lomeli, Lucas Hosseini, Fabio Petroni, Timo Schick, Jane Dwivedi-Yu, Armand Joulin, Sebastian Riedel, and Edouard Grave. Few-shot learning with retrieval augmented language models. arXiv preprint arXiv:2208.03299, 2(3),

    arXiv

  5. [19]

    Learning to retrieve prompts for in-context learning

    22 arXiv Template A PREPRINT Ohad Rubin, Jonathan Herzig, and Jonathan Berant. Learning to retrieve prompts for in-context learning. arXiv preprint arXiv:2112.08633,

    arXiv

  6. [20]

    A survey on in-context learning

    Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Zhiyong Wu, Baobao Chang, Xu Sun, Jingjing Xu, and Zhifang Sui. A survey on in-context learning. arXiv preprint arXiv:2301.00234, 2022a. Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Jingyuan Ma, Rui Li, Heming Xia, Jingjing Xu, Zhiyong ...

    arXiv

  7. [21]

    Benchmarking large language models in retrieval-augmented generation (2023)

    Jiawei Chen, Hongyu Lin, Xianpei Han, and Le Sun. Benchmarking large language models in retrieval-augmented generation (2023). arXiv preprint arXiv:2309.01431,

    2023 arXiv

  8. [22]

    Huggingface’s transformers: State-of-the-art natural language processing

    Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pierric Cistac, Tim Rault, Rémi Louf, Morgan Funtowicz, et al. Huggingface’s transformers: State-of-the-art natural language processing. arXiv preprint arXiv:1910.03771,

    1910 arXiv

  9. [23]

    SPLADE v2: Sparse lexical and expansion model for information retrieval

    Thibault Formal, Carlos Lassance, Benjamin Piwowarski, and Stéphane Clinchant. SPLADE v2: Sparse lexical and expansion model for information retrieval. CoRR, abs/2109.10086,

    arXiv

  10. [24]

    Mujeen Sung, Hwisang Jeon, Jinhyuk Lee, and Jaewoo Kang

    doi:10.1093/bib/bbab282. Mujeen Sung, Hwisang Jeon, Jinhyuk Lee, and Jaewoo Kang. Biomedical entity representations with synonym marginalization. In Dan Jurafsky, Joyce Chai, Natalie Schluter, and Joel Tetreault, editors, Proceedings of the 58th Annual Meeting of the Associati...

    doi

  11. [26]

    Tiantian Zhu, Yang Qin, Ming Feng, Qingcai Chen, Baotian Hu, and Yang Xiang

    doi:10.24963/ijcai.2022/560. Tiantian Zhu, Yang Qin, Ming Feng, Qingcai Chen, Baotian Hu, and Yang Xiang. Biopro: Context-infused prompt learning for biomedical entity linking. IEEE/ACM Transactions on Audio, Speech, and Language Processing, 32: 374–385,

    2022 doi

  12. [27]

    Gpt-4 technical report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774,

    arXiv

  13. [28]

    Llama: Open and efficient foundation language models

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971,

    arXiv

  14. [29]

    Entity matching using large language models

    Ralph Peeters, Aaron Steiner, and Christian Bizer. Entity matching using large language models. arXiv preprint arXiv:2310.11244,

    arXiv

  15. [30]

    Exploring the in-context learning ability of large language model for biomedical concept linking

    Qinyong Wang, Zhenxiang Gao, and Rong Xu. Exploring the in-context learning ability of large language model for biomedical concept linking. ArXiv, abs/2307.01137,

    arXiv

  16. [31]

    Jixiong Liu, Yoan Chabot, Raphaël Troncy, Viet-Phi Huynh, Thomas Labbé, and Pierre Monnin

    doi:10.48550/arXiv.2307.01137. Jixiong Liu, Yoan Chabot, Raphaël Troncy, Viet-Phi Huynh, Thomas Labbé, and Pierre Monnin. From tabular data to knowledge graphs: A survey of semantic table interpretation tasks and methods. Journal of Web Semantics, 76: 100761,

    doi arXiv

  17. [32]

    Johnson, D

    Jiao Li, Yueping Sun, Robin J. Johnson, D. Sciaky, Chih-Hsuan Wei, Robert Leaman, A. P. Davis, C. Mattingly, Thomas C. Wiegers, and Zhiyong Lu. Biocreative v cdr task corpus: a resource for chemical disease relation extraction. Database: The Journal of Biological Databases and...

    2016

  18. [34]

    doi:10.1007/s12471-019-01356-2

    ISSN 1568-5888. doi:10.1007/s12471-019-01356-2. M.L. Handoko participates in the MONITOR-HF trial. A.A. van de Bovenkamp declares no competing interests. Mujeen Sung, Hwisang Jeon, Jinhyuk Lee, and Jaewoo Kang. Biomedical entity representations with synonym marginalization. ar...

    arXiv

  19. [2010]

    Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, et al

    doi:10.1136/jamia.2009.001560. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, et al. A survey of large language models. arXiv preprint arXiv:2303.18223,

    2009 arXiv

  20. [2014]

    Jinhyuk Lee, Wonjin Yoon, Sungdong Kim, Donghyeon Kim, Sunkyu Kim, Chan Ho So, and Jaewoo Kang

    doi:10.1186/1472-6947- 15-S1-S4. Jinhyuk Lee, Wonjin Yoon, Sungdong Kim, Donghyeon Kim, Sunkyu Kim, Chan Ho So, and Jaewoo Kang. Biobert: a pre-trained biomedical language representation model for biomedical text mining. Bioinformatics, 36(4):1234–1240,

    doi

  21. [2016]

    Rezarta Islamaj Do˘gan, Robert Leaman, and Zhiyong Lu

    doi:10.1093/database/baw068. Rezarta Islamaj Do˘gan, Robert Leaman, and Zhiyong Lu. NCBI disease corpus: A resource for disease name recognition and concept normalization. J. Biomed. Inform., 47:1–10, February

    doi

  22. [2020]

    Abigail E Whitlock, Gondy Leroy, Fariba M Donovan, and John N Galgiani

    doi:10.1186/s13643-020-01485-5. Abigail E Whitlock, Gondy Leroy, Fariba M Donovan, and John N Galgiani. Icd codes are insufficient to create datasets for machine learning: An evaluation using all of us data for coccidioidomycosis and myocardial infarction. In 2024 IEEE 12th In...

    doi

  23. [2021]

    Zekeriya Anil Guven and Andre Lamurias

    doi:10.1109/TKDE.2021.3117715. Zekeriya Anil Guven and Andre Lamurias. Multilingual bi-encoder models for biomedical entity linking. Expert Systems, 40,

    2021

  24. [2022]

    Evan French and Bridget T McInnes

    doi:10.1136/bmjhci-2022-100633. Evan French and Bridget T McInnes. An overview of biomedical entity linking throughout the years. Journal of biomedical informatics, 137:104252,

    2022 doi

  25. [2023]

    Vincenzo Cutrona, Federico Bianchi, Ernesto Jiménez-Ruiz, and Matteo Palmonari

    doi:10.1111/exsy.13388. Vincenzo Cutrona, Federico Bianchi, Ernesto Jiménez-Ruiz, and Matteo Palmonari. Tough tables: Carefully evaluating entity linking for tabular data. In International Semantic Web Conference, pages 328–343. Springer,

    doi

  26. [2024]

    Lewis, Wen tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela

    Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Kuttler, M. Lewis, Wen tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. Retrieval-augmented generation for knowledge-intensive nlp tasks. ArXiv, abs/2005.11401,

    2005 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.