Percorrer por autor "Rodrigues, Ricardo"
A mostrar 1 - 10 de 28
Resultados por página
Opções de ordenação
- AIA-BDE: A Corpus of FAQs in Portuguese and their VariationsPublication . Oliveira, Hugo Gonçalo; Ferreira, João; Santos, José; Fialho, Pedro; Rodrigues, Ricardo; Coheur, Luísa; Alves, AnaWe present AIA-BDE, a corpus of 380 domain-oriented FAQs in Portuguese and their variations, i.e., paraphrases or entailed questions, created manually, by humans, or automatically, with Google Translate. Its aims to be used as a benchmark for FAQ retrieval and automatic question-answering, but may be useful in other contexts, such as the development of task-oriented dialogue systems, or models for natural language inference in an interrogative context. We also report on two experiments. Matching variations with their original questions was not trivial with a set of unsupervised baselines, especially for manually created variations. Besides high performances obtained with ELMo and BERT embeddings, an Information Retrieval system was surprisingly competitive when considering only the first hit. In the second experiment, text classifiers were trained with the original questions, and tested when assigning each variation to one of three possible sources, or assigning them as out-of-domain. Here, the difference between manual and automatic variations was not so significant.
- ASAPP 2.0: Advancing the state-of-the-art of semantic textual similarity for PortuguesePublication . Alves, Ana; Oliveira, Hugo Gonçalo; Rodrigues, Ricardo; Encarnação, RuiSemantic Textual Similarity (STS) aims at computing the proximity of meaning transmitted by two sentences. In 2016, the ASSIN shared task targeted STS in Portuguese and released training and test collections. This paper describes the development of ASAPP, a system that participated in ASSIN, but has been improved since then, and now achieves the best results in this task. ASAPP learns a STS function from a broad range of lexical, syntactic, semantic and distributional features. This paper describes the features used in the current version of ASAPP, and how they are exploited in a regression algorithm to achieve the best published results for ASSIN to date, in both European and Brazilian Portuguese.
- ASAPP: alinhamento semântico automático de palavras aplicado ao portuguêsPublication . Alves, Ana Oliveira; Rodrigues, RicardoApresentamos duas abordagens distintas `a tarefa de avalia¸c˜ao conjunta ASSIN onde, dada uma cole¸c˜ao de pares de frases escritas em portuguˆes, s˜ao colocados dois objectivos para cada par: (a) calcular a similaridade semˆantica entre as duas frases; e (b) verificar se uma frase do par ´e par´afrase ou inferˆencia da outra. Uma primeira abordagem, apelidada de Reciclagem, baseia-se exclusivamente em heur´ısticas sobre redes semˆanticas para a l´ıngua portuguesa. A segunda abordagem, apelidada de ASAPP, baseia-se em aprendizagem autom´atica supervisionada. Acima de tudo, os resultados da abordagem Reciclagem permitem comparar, de forma indireta, um conjunto de redes semˆanticas, atrav´es do seu desempenho nesta tarefa. Estes resultados, algo modestos, foram depois utilizados como caracter´ısticas da abordagem ASAPP, juntamente com caracter´ısticas adicionais, ao n´ıvel lexical e sint´atico. Ap´os compara¸c˜ao com os resultados da cole¸c˜ao dourada, verifica-se que a abordagem ASAPP supera a abordagem Reciclagem de forma consistente. Isto ocorre tanto para o Portuguˆes Europeu como para o Portuguˆes Brasileiro, onde o desempenho atinge uma exatid˜ao de 80.28%±0.019 para a inferˆencia textual, enquanto que a correla¸c˜ao dos valores atribu´ıdos para a similaridade semˆantica com aqueles atribu´ıdos por humanos ´e de 66.5% ± 0.021.
- Assessing Factoid Question-Answer Generation for PortuguesePublication . Ferreira, João; Rodrigues, Ricardo; Oliveira, Hugo GonçaloWe present work on the automatic generation of question-answer pairs in Portuguese, useful, for instance, for populating the knowledge-base of question-answering systems. This includes: (i) a new corpus of close to 600 factoid sentences, manually created from an existing corpus of questions and answers, used as our benchmark; (ii) two approaches for the automatic generation of question-answer pairs, which can be seen as baselines; (iii) results of those approaches in the corpus.
- Assessment of perioperative mortality risk in patients with infective endocarditis undergoing cardiac surgery: performance of the EuroSCORE I and II logistic modelsPublication . Madeira, Sérgio; Rodrigues, Ricardo; Tralhão, António; Santos, Miguel; Almeida, Carla; Marques, Marta; Ferreira, Jorge; Raposo, Luís; Neves, José; Mendes, MiguelOBJECTIVES: The European System for Cardiac Operative Risk Evaluation (EuroSCORE) has been established as a tool for assisting decision- making in surgical patients and as a benchmark for quality assessment. Infective endocarditis often equires surgical treatment and is associated with high mortality. This study was undertaken to (i) validate both versions of the EuroSCORE, the older logistic EuroSCORE I and the recently developed EuroSCORE II and to compare their performances; (ii) identify predictors other than those included in the EuroSCORE models that might further improve their performance. METHODS: We retrospectively studied 128 patients from a single-centre registry who underwent heart surgery for active infective endocarditis between January 2007 and November 2014. Binary logistic regression was used to find independent predictors of mortality and to create a new prediction model. Discrimination and calibration of models were assessed by receiver-operating characteristic curve analysis, calibration curves and the Hosmer–Lemeshow test. RESULTS: The observed perioperative mortality was 16.4% (n = 21). The median EuroSCORE I and EuroSCORE II were 13.9% interquartile range (IQ) (7.0–35.0) and 6.6% IQ (3.5–18.2), respectively. Discriminative power was numerically higher for EuroSCORE II {area under the curve (AUC) of 0.83 [95% confidence interval (CI), 0.75–0.91]} than for EuroSCORE I [0.75 (95% CI, 0.66–0.85), P = 0.09]. The Hosmer– Lemeshow test showed good calibration for EuroSCORE II (P = 0.08) but not for EuroSCORE I (P = 0.04). EuroSCORE I tended to overpredict and EuroSCORE II to under-predict mortality. Among the variables known to be associated with greater infective endocarditis severity, only prosthetic valve infective endocarditis remained an independent predictor of mortality [odds ratio (OR) 6.6; 95% CI, 1.1–39.5; P = 0.04]. The new model including the EuroSCORE II variables and variables known to be associated with greater infective endocarditis severity showed an AUC of 0.87 (95% CI, 0.79–0.94) and differed significantly from EuroSCORE I (P = 0.03) but not from EuroSCORE II (P = 0.4). CONCLUSIONS: Both EuroSCORE I and II satisfactorily stratify risk in active infective endocarditis; however, EuroSCORE II performed better in the overall comparison. Specific endocarditis features will increase model complexity without an unequivocal improvement in predictive ability.
- Can We Assess the Success of Surgery for Degenerative Spinal Diseases Using Patients' Recall of Their Preoperative Status?Publication . Rodrigues, Ricardo; Silva, Pedro Santos; Cunha, Marisa; Vaz, Rui; Pereira, PauloBackground: Patients' recall of their preoperative status is seldom used to assess surgical outcomes because of concerns of inaccuracy and bias. Objective: The present study aims to measure the significance of this recall bias and its repercussion on patients' recollection of their preoperative status. Methods: Patients submitted to surgery due to degenerative spine diseases during the period of one year (n=198) were included in this study. EQ-5D (including EQ VAS), COMI Neck (including Neck Pain and Shoulder/Arm Pain NRS), COMI Back (including Back Pain and Buttock/Leg Pain NRS), NDI and ODI were completed preoperatively. One year after surgery, patients were asked to complete 2 sets of the same questionnaires, one regarding their postoperative status and the other one regarding their recall of the preoperative status. Results: There was poor to moderate agreement between recalled and collected preoperative scores for all PROMs. Patients' recollection of their preoperative status was accurate for patients who underwent cervical spine surgery, but not after lumbar spine surgery. Patients satisfied with the outcome after lumbar spine surgery recalled significantly worse scores compared to the preoperatively collected. Conclusions: Using patients' recall of their preoperative status may lead to an overestimation of the surgery effectiveness, particularly for lumbar spine surgery. The self-assessed surgery effectiveness interferes with the recollection of the baseline status.
- Evaluación del nuevo Contador Hematológico Unicel ® DXH 800 – Beckman Coulter en el estudio citológico de los liquidos biológicosPublication . Reis, Ana Batalha; Matos, Susana; Meireles, Flora; Pinto, Tânia; Limão, Luísa; Silva, Patrícia; Antunes, Cláudia; Amaro, Sílvia; Simões, Catarina; Silva, Cândido; Azevedo, Ana Paula; Rodrigues, Ricardo; Sydor, Yuliya; Viana, João; Costa, Telmo; Correia Júnior, Esmeraldina
- Genetic information improves the prediction of major adverse cardiovascular events in the GENEMACOR populationPublication . Mendonca, Maria Isabel; Henriques, Eva; Borges, Sofia; Sousa, Ana Célia; Pereira, Andreia; Santos, Marina; Temtem, Margarida; Freitas, Sónia; Monteiro, Joel; Sousa, João Adriano; Rodrigues, Ricardo; Guerra, Graça; Palma Reis, RobertoThe inclusion of a genetic risk score (GRS) can modify the risk prediction of coronary artery disease (CAD), providing an advantage over the use of traditional models. The predictive value of the genetic information on the recurrence of major adverse cardiovascular events (MACE) remains controversial. A total of 33 genetic variants previously associated with CAD were genotyped in 1587 CAD patients from the GENEMACOR study. Of these, 18 variants presented an hazard ratio >1, so they were selected to construct a weighted GRS (wGRS). MACE discrimination and reclassification were evaluated by C-Statistic, Net Reclassification Index and Integrated Discrimination Improvement methodologies. After the addition of wGRS to traditional predictors, the C-index increased from 0.566 to 0.572 (p=0.0003). Subsequently, adding wGRS to traditional plus clinical risk factors, this model slightly improved from 0.620 to 0.622 but with statistical significance (p=0.004). NRI showed that 17.9% of the cohort was better reclassified when the primary model was associated with wGRS. The Kaplan-Meier estimator showed that, at 15-year follow-up, the group with a higher number of risk alleles had a significantly higher MACE occurrence (p=0.011). In CAD patients, wGRS improved MACE risk prediction, discrimination and reclassification over the conventional factors, providing better cost-effective therapeutic strategies.
- Green Shipping Effect on Sustainable Economy and Environmental Performance.Publication . Felício, J. Augusto; Rodrigues, Ricardo; Caldeirinha, VitorThis paper focuses on green shipping and its’ influence on the sustainable economy and environmental performance. Based on the green shipping approach, this empirical study examines a survey sample of 193 responses from Portuguese and Spanish executive managers and uses exploratory factor analysis and structural equation model. The Green shipping approach supports the green theory. The results show the importance of green efficiency, green management, and pollution impact. The confirmation of the sizeable influence of green shipping on the sustainable economy and environmental performance constructs constitutes a major contribution to the literature. Green management and green efficiency contribute to controlling the impact of pollution with practical effects on economic sustainability. Another contribution arises from the fact that tax and financial incentives and environmental sustainability regulations indicate the relevance of the pollution impact and sustainable economy.
- Improving NLTK for Processing PortuguesePublication . Ferreira, João; Oliveira, Hugo Gonçalo; Rodrigues, RicardoPython has a growing community of users, especially in the AI and ML fields. Yet, Computational Processing of Portuguese in this programming language is limited, in both available tools and results. This paper describes NLPyPort, a NLP pipeline in Python, primarily based on NLTK, and focused on Portuguese. It is mostly assembled from pre-existent resources or their adaptations, but improves over the performance of existing alternatives in Python, namely in the tasks of tokenization, PoS tagging, lemmatization and NER.
- «
- 1 (current)
- 2
- 3
- »