{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T17:49:29Z","timestamp":1782841769251,"version":"3.54.5"},"reference-count":73,"publisher":"JMIR Publications Inc.","issue":"8","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Med Inform"],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>How to treat a disease remains to be the most common type of clinical question. Obtaining evidence-based answers from biomedical literature is difficult. Analogical reasoning with embeddings from deep learning (embedding analogies) may extract such biomedical facts, although the state-of-the-art focuses on pair-based proportional (pairwise) analogies such as man:woman::king:queen (\u201cqueen = \u2212man +king +woman\u201d).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>This study aimed to systematically extract disease treatment statements with a Semantic Deep Learning (SemDeep) approach underpinned by prior knowledge and another type of 4-term analogy (other than pairwise).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>As preliminaries, we investigated Continuous Bag-of-Words (CBOW) embedding analogies in a common-English corpus with five lines of text and observed a type of 4-term analogy (not pairwise) applying the 3CosAdd formula and relating the semantic fields person and death: \u201cdagger = \u2212Romeo +die +died\u201d (search query: \u2212Romeo +die +died). Our SemDeep approach worked with pre-existing items of knowledge (what is known) to make inferences sanctioned by a 4-term analogy (search query \u2212x +z1 +z2) from CBOW and Skip-gram embeddings created with a PubMed systematic reviews subset (PMSB dataset). Stage1: Knowledge acquisition. Obtaining a set of terms, candidate y, from embeddings using vector arithmetic. Some n-gram pairs from the cosine and validated with evidence (prior knowledge) are the input for the 3cosAdd, seeking a type of 4-term analogy relating the semantic fields disease and treatment. Stage 2: Knowledge organization. Identification of candidates sanctioned by the analogy belonging to the semantic field treatment and mapping these candidates to unified medical language system Metathesaurus concepts with MetaMap. A concept pair is a brief disease treatment statement (biomedical fact). Stage 3: Knowledge validation. An evidence-based evaluation followed by human validation of biomedical facts potentially useful for clinicians.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>We obtained 5352 n-gram pairs from 446 search queries by applying the 3CosAdd. The microaveraging performance of MetaMap for candidate y belonging to the semantic field treatment was F-measure=80.00% (precision=77.00%, recall=83.25%). We developed an empirical heuristic with some predictive power for clinical winners, that is, search queries bringing candidate y with evidence of a therapeutic intent for target disease x. The search queries -asthma +inhaled_corticosteroids +inhaled_corticosteroid and -epilepsy +valproate +antiepileptic_drug were clinical winners, finding eight evidence-based beneficial treatments.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>Extracting treatments with therapeutic intent by analogical reasoning from embeddings (423K n-grams from the PMSB dataset) is an ambitious goal. Our SemDeep approach is knowledge-based, underpinned by embedding analogies that exploit prior knowledge. Biomedical facts from embedding analogies (4-term type, not pairwise) are potentially useful for clinicians. The heuristic offers a practical way to discover beneficial treatments for well-known diseases. Learning from deep learning models does not require a massive amount of data. Embedding analogies are not limited to pairwise analogies; hence, analogical reasoning with embeddings is underexploited.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/16948","type":"journal-article","created":{"date-parts":[[2020,5,4]],"date-time":"2020-05-04T15:54:38Z","timestamp":1588607678000},"page":"e16948","source":"Crossref","is-referenced-by-count":1,"title":["Semantic Deep Learning: Prior Knowledge and a Type of Four-Term Embedding Analogy to Acquire Treatments for Well-Known Diseases"],"prefix":"10.2196","volume":"8","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9469-2068","authenticated-orcid":false,"given":"Mercedes","family":"Arguello Casteleiro","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7121-4450","authenticated-orcid":false,"given":"Julio","family":"Des Diz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0349-7716","authenticated-orcid":false,"given":"Nava","family":"Maroto","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2312-1414","authenticated-orcid":false,"given":"Maria Jesus","family":"Fernandez Prieto","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7317-9182","authenticated-orcid":false,"given":"Simon","family":"Peters","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6538-8187","authenticated-orcid":false,"given":"Chris","family":"Wroe","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8689-3580","authenticated-orcid":false,"given":"Carlos","family":"Sevillano Torrado","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2678-3360","authenticated-orcid":false,"given":"Diego","family":"Maseda Fernandez","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6038-9025","authenticated-orcid":false,"given":"Robert","family":"Stevens","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2020,8,6]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.5455\/aim.2008.16.219-225"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMp1714987"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1814779116"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"ref5","unstructured":"JurafskyDMartinJStanford University20192020-06-01Speech and Language Processinghttps:\/\/web.stanford.edu\/~jurafsky\/slp3\/"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1037\/\/0003-066x.52.1.45"},{"key":"ref7","unstructured":"ACL Anthology2020-06-01Analogy (State of the Art)https:\/\/aclweb.org\/aclwiki\/Analogy_(State_of_the_art)"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/d15-1036"},{"key":"ref9","unstructured":"MikolovTYihWZweigGLinguistic Regularities in Continuous Space Word RepresentationsProceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies2013NAACL'13June 9-14, 2013Atlanta, Georgia, US74651"},{"key":"ref10","unstructured":"DrozdAGladkovaAMatsuokaSWord Embeddings, Analogies, and Machine Learning: Beyond King - Man + Woman = QueenProceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers2016COLING'16December 11-16, 2016Osaka, Japan351930"},{"key":"ref11","first-page":"-","author":"Liu, H","year":"2017","journal-title":"Arxiv"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.5815\/ijmsc.2017.02.01"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.2196\/12310"},{"key":"ref14","unstructured":"BMJ Best Practice20182020-06-01Sepsis in Adultshttp:\/\/bestpractice.bmj.com\/topics\/en-gb\/245"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1007\/s11739-012-0802-0"},{"key":"ref16","unstructured":"National Library of Medicine - National Institutes of Health - NIH2020-06-01Search Strategy Used to Create the PubMed Systematic Reviews Filterhttps:\/\/www.nlm.nih.gov\/bsd\/pubmed_subsets\/sysreviews_strategy.html"},{"key":"ref17","unstructured":"Special Issue on Semantic Deep Learning2020-06-01http:\/\/www.semantic-web-journal.net\/content\/special-issue-semantic-deep-learning"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/1.4.398"},{"key":"ref19","unstructured":"BarthaPStanford Encyclopedia of Philosophy20192020-06-01Analogy and Analogical Reasoninghttps:\/\/plato.stanford.edu\/archives\/spr2019\/entries\/reasoning-analogy\/"},{"key":"ref20","unstructured":"MikolovTChenKCorradoGDeanJEfficient Estimation of Word Representations in Vector SpaceProceedings of 1st International Conference on Learning Representations2013ICLR'13May 2-4, 2013Scottsdale, Arizona, US"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/w14-1618"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1016\/s0378-2166(99)00042-9"},{"key":"ref23","unstructured":"SNOMED Confluence2020-06-01Technical Implementation Guidehttp:\/\/snomed.org\/tig"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1016\/j.apmr.2013.05.032"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1162\/coli_a_00237"},{"key":"ref26","unstructured":"NCBO BioPortal20202020-06-01Semanticscience Integrated Ontologyhttps:\/\/bioportal.bioontology.org\/ontologies\/SIO"},{"key":"ref27","unstructured":"ChenDPetersonJGriffithsTEvaluating Vector-Space Models of AnalogyProceedings of the 39th Annual Meeting of the Cognitive Science Society2017CogSci'17July 26-29, 2017London, UK"},{"key":"ref28","unstructured":"Google Code2020-06-01Word2vec - Defaulthttps:\/\/code.google.com\/archive\/p\/word2vec\/source\/default\/source"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1186\/s13326-019-0212-6"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkh061"},{"key":"ref31","unstructured":"BMJ Best Practice2020-06-01https:\/\/bestpractice.bmj.com\/"},{"key":"ref32","unstructured":"PubMed\/MEDLINE2020-06-10https:\/\/www.ncbi.nlm.nih.gov\/pubmed\/"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.2196\/jmir.5234"},{"key":"ref34","unstructured":"BMJ Best Practice2020-06-01Do You Work for the NHS in England, Scotland or Wales?https:\/\/bestpractice.bmj.com\/info\/bmagp\/"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1037\/0033-295X.104.2.211"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1162\/jmlr.2003.3.4-5.993"},{"key":"ref37","unstructured":"ThomoAEngineering and Computer Science - University of Victoria2020-06-01Latent Semantic Analysishttps:\/\/www.engr.uvic.ca\/~seng474\/svd.pdf"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1016\/j.pragma.2011.03.014"},{"key":"ref39","unstructured":"Cochrane Library: Cochrane Reviews2020-06-01Cochrane Database of Systematic Reviewshttps:\/\/www.cochranelibrary.com\/cdsr\/about-cdsr"},{"key":"ref40","year":"2012","journal-title":"Prevention and Control of Noncommunicable Diseases: Guidelines for Primary Health Care in Low Resource Settings"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1186\/s13326-019-0212-6"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1186\/s13326-018-0181-1"},{"key":"ref43","author":"Manning, C","year":"1999","journal-title":"Foundations of Statistical Natural Language Processing"},{"key":"ref44","unstructured":"NovakJCa\u00f1asACMap20082020-06-01The Theory Underlying Concept Maps and How to Construct and Use Themhttp:\/\/cmap.ihmc.us\/Publications\/ResearchPapers\/TheoryUnderlyingConceptMaps.pdf"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1136\/amiajnl-2011-000464"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1093\/database\/bar013"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1162\/coli.07-034-r2"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.2307\/2288384"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1038\/nrg3337"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1136\/jamia.2009.002733"},{"key":"ref51","unstructured":"Guidelines Developed for Step 4: Named Entity Recognition Task2020-06-01https:\/\/static-content.springer.com\/esm\/art%3A10.1186%2Fs13326-019-0212-6\/MediaObjects\/13326_2019_212_MOESM2_ESM.pdf"},{"key":"ref52","first-page":"529","author":"Pratt, W","year":"2003","journal-title":"AMIA Annu Symp Proc"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.312.7023.71"},{"key":"ref54","doi-asserted-by":"crossref","DOI":"10.1075\/z.44","author":"Sager, J","year":"1990","journal-title":"A Practical Course in Terminology Processing"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1177\/001316446002000104"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1177\/1751143718765407"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1189\/jlb.5MR0114-067R"},{"key":"ref58","unstructured":"NCBI Bookshelf20192020-06-10UMLS Reference Manualhttps:\/\/www.ncbi.nlm.nih.gov\/books\/NBK9685\/table\/ch03.T.related_concepts_file_mrrel_rrf\/"},{"issue":"5","key":"ref59","first-page":"360","volume":"37","author":"Viera, AJ","year":"2005","journal-title":"Fam Med"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1016\/0895-4356(90)90158-l"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1016\/0895-4356(90)90159-m"},{"key":"ref62","author":"Lehrer, A","year":"1985","journal-title":"The Influence of Semantic Fields on Semantic Change"},{"key":"ref63","unstructured":"BMJ Best Practice2020-06-01Asthma in Adultshttp:\/\/bestpractice.bmj.com\/topics\/en-gb\/44"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00134"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/w16-2922"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1201\/9780429469275-13"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btw529"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2016.2615052"},{"key":"ref69","unstructured":"Government of UK2020-06-01Personalised Health and Care 2020https:\/\/www.gov.uk\/government\/publications\/personalised-health-and-care-2020"},{"key":"ref70","unstructured":"National Library of Medicine - National Institutes of Health - NIH2020-06-01The CORE Problem List Subset of SNOMED CThttps:\/\/www.nlm.nih.gov\/research\/umls\/Snomed\/core_subset.html"},{"key":"ref71","unstructured":"NHS Digital2020-06-01SNOMED CT Subset Members in Spreadsheet Viewhttps:\/\/isd.digital.nhs.uk\/trud3\/user\/guest\/group\/0\/pack\/40"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1016\/j.jbi.2006.06.004"},{"key":"ref73","first-page":"572","volume":"2010","author":"Pakhomov, S","year":"2010","journal-title":"AMIA Annu Symp Proc"}],"container-title":["JMIR Medical Informatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/medinform.jmir.org\/2020\/8\/e16948","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T17:19:26Z","timestamp":1782839966000},"score":1,"resource":{"primary":{"URL":"https:\/\/medinform.jmir.org\/2020\/8\/e16948"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,8,6]]},"references-count":73,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2020]]}},"URL":"https:\/\/doi.org\/10.2196\/16948","relation":{"has-preprint":[{"id-type":"doi","id":"10.2196\/preprints.16948","asserted-by":"object"}]},"ISSN":["2291-9694"],"issn-type":[{"value":"2291-9694","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,8,6]]}}}