{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T18:09:11Z","timestamp":1792692551174,"version":"4.5.0"},"reference-count":55,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J Med Internet Res"],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Social media sites are becoming an increasingly important source of information about mental health disorders. Among them, eating disorders are complex psychological problems that involve unhealthy eating habits. In particular, there is evidence showing that signs and symptoms of anorexia nervosa can be traced in social media platforms. Knowing that input data biases tend to be amplified by artificial intelligence algorithms and, in particular, machine learning, these methods should be revised to mitigate biased discrimination in such important domains.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>The main goal of this study was to detect and analyze the performance disparities across genders in algorithms trained for the detection of anorexia nervosa on social media posts. We used a collection of automated predictors trained on a data set in Spanish containing cases of 177 users that showed signs of anorexia (471,262 tweets) and 326 control cases (910,967 tweets).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We first inspected the predictive performance differences between the algorithms for male and female users. Once biases were detected, we applied a feature-level bias characterization to evaluate the source of such biases and performed a comparative analysis of such features and those that are relevant for clinicians. Finally, we showcased different bias mitigation strategies to develop fairer automated classifiers, particularly for risk assessment in sensitive domains.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Our results revealed concerning predictive performance differences, with substantially higher false negative rates (FNRs) for female samples (FNR=0.082) compared with male samples (FNR=0.005). The findings show that biological processes and suicide risk factors were relevant for classifying positive male cases, whereas age, emotions, and personal concerns were more relevant for female cases. We also proposed techniques for bias mitigation, and we could see that, even though disparities can be mitigated, they cannot be eliminated.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>We concluded that more attention should be paid to the assessment of biases in automated methods dedicated to the detection of mental health issues. This is particularly relevant before the deployment of systems that are thought to assist clinicians, especially considering that the outputs of such systems can have an impact on the diagnosis of people at risk.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/45184","type":"journal-article","created":{"date-parts":[[2023,6,8]],"date-time":"2023-06-08T10:45:34Z","timestamp":1686221134000},"page":"e45184","source":"Crossref","is-referenced-by-count":21,"title":["Gender Bias When Using Artificial Intelligence to Assess Anorexia Nervosa on Social Media: Data-Driven Study"],"prefix":"10.2196","volume":"25","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6979-9330","authenticated-orcid":false,"given":"David","family":"Solans Noguero","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0742-4773","authenticated-orcid":false,"given":"Diana","family":"Ram\u00edrez-Cifuentes","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7072-4096","authenticated-orcid":false,"given":"Esteban Andr\u00e9s","family":"R\u00edssola","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4698-2129","authenticated-orcid":false,"given":"Ana","family":"Freire","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2023,6,8]]},"reference":[{"key":"ref1","unstructured":"LiuMYBreuelTKautzJUnsupervised image-to-image translation networksProceedings of the 31st International Conference on Neural Information Processing Systems2017NIPS '17December 4-9, 2017Long Beach, CA, USA7008"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-02267-3"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/msp.2012.2205597"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-018-9654-y"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.2196\/25925"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1016\/j.ipm.2022.102890"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-28577-7 _27"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/w19-3003"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.2196\/17758"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1126\/science.aal4230"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3457607"},{"key":"ref12","unstructured":"AngwinJLarsonLMattuSKirchnerLMachine bias: there's software used across the country to predict future criminals. And it's biased against BlacksPropublica201605232023-03-01https:\/\/www.propublica.org\/article\/machine-bias-risk-assess ments-in-criminal-sentencing"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/icde.2019.00121"},{"key":"ref14","unstructured":"BuolamwiniJGebruTGender shades: intersectional accuracy disparities in commercial gender classificationProceedings of the 1st Conference on Fairness, Accountability and Transparency2018PMLR '18February 23-24, 2018New York, NY, USA"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1093\/jamiaopen\/ooz054"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1016\/j.amepre.2017.09.008"},{"issue":"7093","key":"ref17","doi-asserted-by":"crossref","first-page":"1536","DOI":"10.1136\/bmj.314.7093.1536","volume":"314","author":"Davies, T","year":"1997","journal-title":"BMJ"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3422824"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-020-0233-7"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1145\/3437259"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0086191"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/w14-3207"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1145\/3372278.3391932"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.2196\/26769"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/tkde.2017.2786695"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3287560.3287587"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0240376"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.eacl-main.256"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/w15-1204"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/w17-1612"},{"key":"ref31","author":"Prochaska, JO","year":"1997","journal-title":"Health Behavior and Health Education: Theory, Research, and Practice. 2nd edition"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1145\/3308558.3313684"},{"key":"ref33","unstructured":"BarocasSHardtMNarayananAFairness and machine learning: Limitations and opportunitiesNIPS Tutorial20172023-02-01https:\/\/fairmlbook.org\/pdf\/fairmlbook.pdf"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1023\/a:1012487302797"},{"key":"ref35","first-page":"2825","volume":"12","author":"Pedregosa, F","year":"2011","journal-title":"J Mach Learn Res"},{"key":"ref36","unstructured":"CalmonFPWeiDVinzamuriBRamamurthyKNVarshneyKROptimized pre-processing for discrimination preventionProceedings of the 31st International Conference on Neural Information Processing Systems2017NIPS '17December 4-9, 2017Red Hook, NY, USA39954004"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1007\/s10115-011-0463-8"},{"key":"ref38","unstructured":"PleissGRaghavanMWuFKleinbergJMWeinbergerKQOn fairness and calibrationProceedings of the 31st International Conference on Neural Information Processing Systems2017NIPS '17December 4-9, 2017Red Hook, NY, USA568493"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1037\/lhb0000206"},{"key":"ref40","unstructured":"KleinbergJMullainathanSRaghavanMInherent trade-offs in the fair determination of risk scoresProceedings of the 8th Innovations in Theoretical Computer Science Conference2017ITCS '17January 9-11, 2017Berkeley, CA, USA123"},{"key":"ref41","unstructured":"Probability calibrationscikit-learn developers2023-03-27https:\/\/scikit-learn.org\/stable\/modules\/calibration.html"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/e17-1015"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-018-0036-2"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1037\/h0031619"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1037\/\/0033-2909.86.2.420"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.20982\/tqmp.08.1.p023"},{"key":"ref47","author":"Roach\u00e9, DJ","year":"2017","journal-title":"The SAGE Encyclopedia of Communication Research Methods"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/w18-0603"},{"issue":"2","key":"ref49","first-page":"58","volume":"8","author":"Freeman, AC","year":"2005","journal-title":"S Afr Psychiatry Rev"},{"key":"ref50","unstructured":"LundbergSMLeeSIA unified approach to interpreting model predictionsProceedings of the 31st International Conference on Neural Information Processing Systems2017NIPS '17December 4-9, 2017Long Beach, CA, USA476877"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1103\/physreve.69.066138"},{"key":"ref52","unstructured":"GiuseppeMDealing with fairness in public policy analysis: a methodological frameworkEuropean Commission20172023-03-01https:\/\/op.europa.eu\/en\/publication-detail\/-\/publication\/94fd85ed-a8b1-11e7-837e-01aa75ed71a1\/language-en"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.3389\/fpsyt.2019.00864"},{"key":"ref54","year":"2013","journal-title":"Diagnostic and Statistical Manual of Mental Disorders (DSM-5), 5th Edition"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.3389\/fdata.2019.00013"}],"container-title":["Journal of Medical Internet Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.jmir.org\/2023\/1\/e45184","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T17:50:28Z","timestamp":1782841828000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.jmir.org\/2023\/1\/e45184"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,8]]},"references-count":55,"URL":"https:\/\/doi.org\/10.2196\/45184","relation":{"has-preprint":[{"id-type":"doi","id":"10.2196\/preprints.45184","asserted-by":"object"}]},"ISSN":["1438-8871"],"issn-type":[{"value":"1438-8871","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,6,8]]}}}