Medikuntza arloko eta arlo juridikoko hizkuntzaren prozesamendua

Osasunaren eta legearen domeinuak

Hizkuntza prozesatzeko teknikak domeinu zehatzetara eramaten direnean, osasunaren eta legearen domeinuetara adibidez, hainbat egokitzapen behar izaten dira testu horien berezitasunak modu egokian jasotzeko.

Osasunaren eta legearen domeinuak

Osasunaren domeinuan Galdakaoko-Usansoloko Ospitalearekin 2010ean hasi ginen elkarlanean Gaixotasunen Nazioarteko Sailkapenaren (GNS) kodeketa automatikoa hobetzen. Orduz geroztik, eta beti ere pazienteari ematen zaion arreta hobetzeko asmoz, hainbat atazetan egin dugu lan: botiken aurkako erreakzioak detektatzen, osasunaren alorreko entitate izendunak ezagutzen (botika-izenak, gaixotasunak, gorputz-egiturak e.a.) eta azken aldian, osasun-txostenen itzulpen automatikoan. Lan gehienak gaztelerazko osasun-txostenekin egin baditugu ere, ingeleseko corpusak ere erabili ditugu, eta guretzat oso garrantzitsua dena, euskara osasun-zerbitzuetara eramateko urrats garrantzitsuak eman ditugu. Ildo honetan, SNOMED CT medikuntzako lexikoi erraldoia automatikoki euskaratu dugu eta euskara eta gaztelera arteko itzulpen automatiko neuronalarekin lehen emaitzak lortu ditugu.

Osakidetzako hainbat ospitalekin batera eskatu ditugu OSAKU, DETEAMI, EXTRECM eta PROSAMED proiektuak, eta Osakidetzako Euskara Zerbitzuaren enkarguz GNS-10 sailkapenaren itzulpen automatikoa egin dugu.

Domainu legalaren alorrean 2018an Minsait enpresarentzat lanean aritu ginen Prótagoras izeneko barne proiektu batean laguntzen. Proiektuak Hizkuntza Naturalaren Ulermena du helburu eta lehen pausu bezala, domeinu legaleko dokumentatze-prozesuan zentratu ginen. Jakin badakigu dokumentu multzo izugarri handiak sortzen direla eta berebiziko garrantzia duela datu andana horien antolaketa eta biltegiratze azkarrak egitea erabakiak era efikazean hartzeko.

Proiektuaren emaitza bezala kasu erreal bateko informazio erauzketa egiteko soluzioa aurkeztu genuen. Notaritzako eskrituretatik informazio erauzteko gai den sistema garatu genuen. Balio erantsi bezala, gure sistemak ez du inolako datu anotaturik erabiltzen patroiak ikasteko, horrela, dokumentu horietatik deskribatutako ondasunen ezaugarri nagusiak (eraikinaren helbidea, balioa eta abar) antzemateaz gain jabeen informazioa (izen-abizenak, jabetza mota, eta abar) modu merkean erauzteko gai da.

Gutxiago irakurri

Ikertzaile nagusia:

Arantza Casillas

ikertzaileak:

Arantza Díaz de Ilarraza

Nerea Ezeiza

Iakes Goenaga

Koldo Gojenola

Oier Lopez de Lacalle

Maite Oronoz

Alicia Pérez

Olatz Perez de Viñaspre

domains_tabs

Demoak

Clinical entity and relation extraction in Spanish

Insert a text from the clinical domain and the systems will detect the disorders, drugs, body parts and procedures in it as well as adverse drug reactions and relations between disorders

Kontratuak

Proiektuak

Patenteak

System to detect Spanish medical entities in the medical domain

Argitalpenak

Alberto Blanco Olatz Perez de Viñaspre Alicia Pérez Arantza Casillas

Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity (2020)

https://www.sciencedirect.com/science/article/pii/S0169260719311745

Rebecka Weegar, Alicia Pérez, Arantza Casillas, Maite Oronoz

Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches (2020)

BMC Medical Informatics and Decision Making

Sara Santiso

Adverse Drug Reaction extraction on Electronic Health Records written in Spanish (2020)

Procesamiento del Lenguaje Natural http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6203

Sara Santiso, Alicia Pérez, Arantza Casillas, Maite Oronoz

Neural negated entity recognition in Spanish electronic health records (2020)

Journal of Biomedical Informatics (JBI) https://doi.org/10.1016/j.jbi.2020.103419

Olatz Perez-de-Viñaspre, Maite Oronoz, Natalia Elvira

KabiTermICD: Nested Term Based Translation of the ICD-10-CM into a Minor Language (2018)

Workshop "MultilingualBIO: Multilingual Biomedical Text Processing" of LREC 2018. Proceedings of the workshop. Miyazaki (Japan), 8th May 2018.

Jorge Pérez, Alicia Pérez, Arantza Casillas, Koldo Gojenola

Cardiology record multi-label classification using Latent Dirichlet Allocation (2018)

Computer Methods and Programs in Biomedicine https://doi.org/10.1016/j.cmpb.2018.07.002

Aitziber Atutxa, Arantza Casillas, Nerea Ezeiza, Iakes Goenaga, V. Fresno, Koldo Gojenola, R. Martinez, Maite Oronoz, Olatz Perez-de-Viñaspre

IxaMed at CLEF eHealth 2018 Task 1: ICD10 Coding with a Sequence-to-Sequence approach (2018)

CLEF 2018 Online Working Notes. CEUR-WS

Mikel Laburu, Alicia Pérez, Arantza Casillas, Iakes Goenaga, Maite Oronoz

Can I find information about rare diseases in some other language? (2018)

IEEE International Conference on Bioinformatics and Biomedicine. Artificial Intelligence techniques for Biomedicine and Healthcare. Madrid (December, 2018); ISBN: 978-1-5386-5487-3; Pgs: 2102-2108

More publications

domains_tabs_full

Clinical entity and relation extraction in Spanish

Insert a text from the clinical domain and the systems will detect the disorders, drugs, body parts and procedures in it as well as adverse drug reactions and relations between disorders

System to detect Spanish medical entities in the medical domain

Alberto Blanco Olatz Perez de Viñaspre Alicia Pérez Arantza Casillas

Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity (2020)

https://www.sciencedirect.com/science/article/pii/S0169260719311745

Rebecka Weegar, Alicia Pérez, Arantza Casillas, Maite Oronoz

Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches (2020)

BMC Medical Informatics and Decision Making

Sara Santiso

Adverse Drug Reaction extraction on Electronic Health Records written in Spanish (2020)

Procesamiento del Lenguaje Natural http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6203

Sara Santiso, Alicia Pérez, Arantza Casillas, Maite Oronoz

Neural negated entity recognition in Spanish electronic health records (2020)

Journal of Biomedical Informatics (JBI) https://doi.org/10.1016/j.jbi.2020.103419

Olatz Perez-de-Viñaspre, Maite Oronoz, Natalia Elvira

KabiTermICD: Nested Term Based Translation of the ICD-10-CM into a Minor Language (2018)

Workshop "MultilingualBIO: Multilingual Biomedical Text Processing" of LREC 2018. Proceedings of the workshop. Miyazaki (Japan), 8th May 2018.

Jorge Pérez, Alicia Pérez, Arantza Casillas, Koldo Gojenola

Cardiology record multi-label classification using Latent Dirichlet Allocation (2018)

Computer Methods and Programs in Biomedicine https://doi.org/10.1016/j.cmpb.2018.07.002

Aitziber Atutxa, Arantza Casillas, Nerea Ezeiza, Iakes Goenaga, V. Fresno, Koldo Gojenola, R. Martinez, Maite Oronoz, Olatz Perez-de-Viñaspre

IxaMed at CLEF eHealth 2018 Task 1: ICD10 Coding with a Sequence-to-Sequence approach (2018)

CLEF 2018 Online Working Notes. CEUR-WS

Mikel Laburu, Alicia Pérez, Arantza Casillas, Iakes Goenaga, Maite Oronoz

Can I find information about rare diseases in some other language? (2018)

IEEE International Conference on Bioinformatics and Biomedicine. Artificial Intelligence techniques for Biomedicine and Healthcare. Madrid (December, 2018); ISBN: 978-1-5386-5487-3; Pgs: 2102-2108

More publications

Languages

Medikuntza arloko eta arlo juridikoko hizkuntzaren prozesamendua

Arantza Casillas

domains_tabs

Demoak

Clinical entity and relation extraction in Spanish

Kontratuak

Proiektuak

Patenteak

System to detect Spanish medical entities in the medical domain

Argitalpenak

Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity (2020)

Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches (2020)

Adverse Drug Reaction extraction on Electronic Health Records written in Spanish (2020)

Neural negated entity recognition in Spanish electronic health records (2020)

KabiTermICD: Nested Term Based Translation of the ICD-10-CM into a Minor Language (2018)

Cardiology record multi-label classification using Latent Dirichlet Allocation (2018)

IxaMed at CLEF eHealth 2018 Task 1: ICD10 Coding with a Sequence-to-Sequence approach (2018)

Can I find information about rare diseases in some other language? (2018)

domains_tabs_full

Clinical entity and relation extraction in Spanish

System to detect Spanish medical entities in the medical domain

Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity (2020)

Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches (2020)

Adverse Drug Reaction extraction on Electronic Health Records written in Spanish (2020)

Neural negated entity recognition in Spanish electronic health records (2020)

KabiTermICD: Nested Term Based Translation of the ICD-10-CM into a Minor Language (2018)

Cardiology record multi-label classification using Latent Dirichlet Allocation (2018)

IxaMed at CLEF eHealth 2018 Task 1: ICD10 Coding with a Sequence-to-Sequence approach (2018)

Can I find information about rare diseases in some other language? (2018)

HiTZ Ikerkuntza talde hauek osatzen dute :

Languages

You are here

Medikuntza arloko eta arlo juridikoko hizkuntzaren prozesamendua

domains_tabs

System to detect Spanish medical entities in the medical domain

Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity (2020)

Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches (2020)

Adverse Drug Reaction extraction on Electronic Health Records written in Spanish (2020)

Neural negated entity recognition in Spanish electronic health records (2020)

KabiTermICD: Nested Term Based Translation of the ICD-10-CM into a Minor Language (2018)

Cardiology record multi-label classification using Latent Dirichlet Allocation (2018)

IxaMed at CLEF eHealth 2018 Task 1: ICD10 Coding with a Sequence-to-Sequence approach (2018)

Can I find information about rare diseases in some other language? (2018)

domains_tabs_full

System to detect Spanish medical entities in the medical domain

Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity (2020)

Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches (2020)

Adverse Drug Reaction extraction on Electronic Health Records written in Spanish (2020)

Neural negated entity recognition in Spanish electronic health records (2020)

KabiTermICD: Nested Term Based Translation of the ICD-10-CM into a Minor Language (2018)

Cardiology record multi-label classification using Latent Dirichlet Allocation (2018)

IxaMed at CLEF eHealth 2018 Task 1: ICD10 Coding with a Sequence-to-Sequence approach (2018)

Can I find information about rare diseases in some other language? (2018)

HiTZ Ikerkuntza talde hauek osatzen dute :