Emergence of the Coronavirus 2019 Disease has highlighted further the need for timely support for... more Emergence of the Coronavirus 2019 Disease has highlighted further the need for timely support for clinicians as they manage severely ill patients. We combine Semantic Web technologies with Deep Learning for Natural Language Processing with the aim of converting human-readable best evidence/practice for COVID-19 into that which is computer-interpretable. We present the results of experiments with 1212 clinical ideas (medical terms and expressions) from two UK national healthcare services specialty guides for COVID-19 and three versions of two BMJ Best Practice documents for COVID-19. The paper seeks to recognise and categorise clinical ideas, performing a Named Entity Recognition (NER) task, with an ontology providing extra terms as context and describing the intended meaning of categories understandable by clinicians. The paper investigates: 1) the performance of classical NER using MetaMap versus NER with fine-tuned BERT models; 2) the integration of both NER approaches using a lig...
This paper has been presented at Open Science International Staff Week celebrated in Getafe (Madr... more This paper has been presented at Open Science International Staff Week celebrated in Getafe (Madrid) - November 6-8, 2019
... Presentada por: María de la Nava Maroto García Dirigida por: ... A D. Ismael Sanz, del Depart... more ... Presentada por: María de la Nava Maroto García Dirigida por: ... A D. Ismael Sanz, del Departamento de Ciencia e Ingeniería de los Computadores de la Universidad Jaume I ya Dª María Jesús José Ferrandis, del Instituto ... Cristina McLaren, Marta Peinado y las chicas de la plaga. ...
Avec des contributions traitant de langues très différentes, cet ouvrage étudie la terminologie à... more Avec des contributions traitant de langues très différentes, cet ouvrage étudie la terminologie à la fois dans les politiques linguistiques, dans les centres de formation et dans les métiers de la traduction pour offrir un vaste panorama des recherches actuelles en la matière.
Additional file 4. This file contains the SPARQL SELECT queries; their results appear in Tables 9... more Additional file 4. This file contains the SPARQL SELECT queries; their results appear in Tables 9 and 11.
Additional file 3. This file shows the results of the evaluation of UMLS CUI pairs with BMJ Best ... more Additional file 3. This file shows the results of the evaluation of UMLS CUI pairs with BMJ Best Practice content (i.e. human medicine), i.e. the file contains the 3-tuples (target concept, candidate concept, validation label) for the VetCN dataset (worksheet "VetCN") and the PMSB dataset (worksheet "PMSB"). The worksheet "signatures" has the ontological signature (i.e. a list of SNOMED CT identifiers) for each of the 11 medical conditions that are the subject of this study. The worksheet "q One Health" shows the number of UMLS CUI pairs validated with BMJ Best Practice content (i.e. human medicine) for each of the 27 UMLS Semantic Types that participates in the SPARQL SELECT query q1VU or q2VU or q3VU (i.e. One Health queries from Table 11).
Additional file 2. This file contains the guidelines developed for "Step 4: Named entity rec... more Additional file 2. This file contains the guidelines developed for "Step 4: Named entity recognition task". The file also contains the section "Avoiding pitfalls from the SemDeep pipeline when extracting locality-based modules with SNOMED CT".
Additional file 1. This file contains 880 term pairs (target term, candidate term) obtained from ... more Additional file 1. This file contains 880 term pairs (target term, candidate term) obtained from the two datasets for 11 medical conditions: the worksheet "VetCN" with the 440 term pairs using CBOW and Skip-gram with the VetCN dataset; and the worksheet "PMSB" with the 440 term pairs using CBOW and Skip-gram with the PMSB dataset. Within the worksheet "VetCN" and "PMSB" appear the UMLS CUIs assigned to the candidate terms (n-grams). The worksheet "SF to LF" has the 63 long forms for 80 short forms (including variants of the short forms) within the candidate terms (n-grams). The worksheet "MetaMap performance" contains the number of TP, FP, and FN obtained and used to calculate precision, recall, and F measure for MetaMap in Experiment 1 (EXP-1) and Experiment 2 (EXP-2).
pairwiseIAA.xls, this file contains the values of the Cohenâ s Kappa measure for each pair of rat... more pairwiseIAA.xls, this file contains the values of the Cohenâ s Kappa measure for each pair of raters per experiment and model, as well as the average mean. (XLS 8Â kb)
TermsMapped_votingSystem.xls, this file contains the details of the voting system for CBOW and Sk... more TermsMapped_votingSystem.xls, this file contains the details of the voting system for CBOW and Skip-gram per experiment. (XLS 70Â kb)
Emergence of the Coronavirus 2019 Disease has highlighted further the need for timely support for... more Emergence of the Coronavirus 2019 Disease has highlighted further the need for timely support for clinicians as they manage severely ill patients. We combine Semantic Web technologies with Deep Learning for Natural Language Processing with the aim of converting human-readable best evidence/practice for COVID-19 into that which is computer-interpretable. We present the results of experiments with 1212 clinical ideas (medical terms and expressions) from two UK national healthcare services specialty guides for COVID-19 and three versions of two BMJ Best Practice documents for COVID-19. The paper seeks to recognise and categorise clinical ideas, performing a Named Entity Recognition (NER) task, with an ontology providing extra terms as context and describing the intended meaning of categories understandable by clinicians. The paper investigates: 1) the performance of classical NER using MetaMap versus NER with fine-tuned BERT models; 2) the integration of both NER approaches using a lig...
This paper has been presented at Open Science International Staff Week celebrated in Getafe (Madr... more This paper has been presented at Open Science International Staff Week celebrated in Getafe (Madrid) - November 6-8, 2019
... Presentada por: María de la Nava Maroto García Dirigida por: ... A D. Ismael Sanz, del Depart... more ... Presentada por: María de la Nava Maroto García Dirigida por: ... A D. Ismael Sanz, del Departamento de Ciencia e Ingeniería de los Computadores de la Universidad Jaume I ya Dª María Jesús José Ferrandis, del Instituto ... Cristina McLaren, Marta Peinado y las chicas de la plaga. ...
Avec des contributions traitant de langues très différentes, cet ouvrage étudie la terminologie à... more Avec des contributions traitant de langues très différentes, cet ouvrage étudie la terminologie à la fois dans les politiques linguistiques, dans les centres de formation et dans les métiers de la traduction pour offrir un vaste panorama des recherches actuelles en la matière.
Additional file 4. This file contains the SPARQL SELECT queries; their results appear in Tables 9... more Additional file 4. This file contains the SPARQL SELECT queries; their results appear in Tables 9 and 11.
Additional file 3. This file shows the results of the evaluation of UMLS CUI pairs with BMJ Best ... more Additional file 3. This file shows the results of the evaluation of UMLS CUI pairs with BMJ Best Practice content (i.e. human medicine), i.e. the file contains the 3-tuples (target concept, candidate concept, validation label) for the VetCN dataset (worksheet "VetCN") and the PMSB dataset (worksheet "PMSB"). The worksheet "signatures" has the ontological signature (i.e. a list of SNOMED CT identifiers) for each of the 11 medical conditions that are the subject of this study. The worksheet "q One Health" shows the number of UMLS CUI pairs validated with BMJ Best Practice content (i.e. human medicine) for each of the 27 UMLS Semantic Types that participates in the SPARQL SELECT query q1VU or q2VU or q3VU (i.e. One Health queries from Table 11).
Additional file 2. This file contains the guidelines developed for "Step 4: Named entity rec... more Additional file 2. This file contains the guidelines developed for "Step 4: Named entity recognition task". The file also contains the section "Avoiding pitfalls from the SemDeep pipeline when extracting locality-based modules with SNOMED CT".
Additional file 1. This file contains 880 term pairs (target term, candidate term) obtained from ... more Additional file 1. This file contains 880 term pairs (target term, candidate term) obtained from the two datasets for 11 medical conditions: the worksheet "VetCN" with the 440 term pairs using CBOW and Skip-gram with the VetCN dataset; and the worksheet "PMSB" with the 440 term pairs using CBOW and Skip-gram with the PMSB dataset. Within the worksheet "VetCN" and "PMSB" appear the UMLS CUIs assigned to the candidate terms (n-grams). The worksheet "SF to LF" has the 63 long forms for 80 short forms (including variants of the short forms) within the candidate terms (n-grams). The worksheet "MetaMap performance" contains the number of TP, FP, and FN obtained and used to calculate precision, recall, and F measure for MetaMap in Experiment 1 (EXP-1) and Experiment 2 (EXP-2).
pairwiseIAA.xls, this file contains the values of the Cohenâ s Kappa measure for each pair of rat... more pairwiseIAA.xls, this file contains the values of the Cohenâ s Kappa measure for each pair of raters per experiment and model, as well as the average mean. (XLS 8Â kb)
TermsMapped_votingSystem.xls, this file contains the details of the voting system for CBOW and Sk... more TermsMapped_votingSystem.xls, this file contains the details of the voting system for CBOW and Skip-gram per experiment. (XLS 70Â kb)
Esta monografía ofrece una visión plural y contrastada sobre distintos aspectos relacionados con ... more Esta monografía ofrece una visión plural y contrastada sobre distintos aspectos relacionados con la creación de neologismos (es, fr, ca, gl, it, pt, ro). Un abanico de propuestas y enfoques formulados por los principales expertos en neología del panorama internacional actual, que permite trazar las líneas de actuación e investigación más relevantes de este ámbito, y que ayuda al lector a comprender las dinámicas de creación de léxico en unas lenguas que, además de un mismo origen, comparten desafíos y oportunidades de cara al futuro. Sin duda, los recursos, estrategias y nuevas orientaciones que se presentan en esta obra arrojan no pocas claves importantes para el estudio y la potenciación de las lenguas románicas en el actual panorama lingüístico global.
Uploads
Papers