RESUMO
Este trabalho tem como objetivo relatar estratégias para coleta de um conjunto de dados em português para treinamento de modelos de Inteligência Artificial com vistas a identificar de forma automática fake news sobre covid-19 disseminadas durante a pandemia, a partir de código Python. Analisamos um método de detecção de fake news baseado em uma Rede Neural Recorrente e de aprendizagem supervisionada. Selecionamos um corpus com 7,2 mil textos coletados em websites e agências de notícias por Monteiro et al. (2018) com cada um previamente catalogado como verdadeiro ou falso como conjunto de dados de treino e validação. O modelo foi usado para detecção de fake news sobre covid-19 em um conjunto de notícias coletadas e classificadas pelos autores deste trabalho. O índice de acerto foi de 70%, ou seja, essa foi a taxa de sucesso da detecção dos itens catalogados.
This work aims to report strategies for collecting a dataset in Portuguese for training Artificial Intelligence models to automatically identify fake news about covid-19 disseminated during the pandemic, using Python code. We analyze a fake news detection method based on a Recurrent Neural Network and supervised learning. We selected a corpus with 7,200 texts collected on websites and news agencies by Monteiro et al. (2018), each one of them previously cataloged as true or false as a training and validation dataset. This model was used to detect fake news about covid-19 in a set of news collected and classified by the authors of this work. The hit rate was 70%.
Este trabajo tiene como objetivo informar estrategias para recopilar un conjunto de datos en portugués para entrenar modelos de Inteligencia Artificial para identificar automáticamente noticias falsas sobre covid-19 difundidas durante la pandemia, utilizando el código Python. Analizamos un método de detección de noticias falsas basado en una Red Neuronal Recurrente y de aprendizaje supervisado. Seleccionamos un corpus de 7.200 textos recogidos en webs y agencias de noticias por Monteiro et al. (2018) con cada uno catalogado previamente como verdadero o falso como un conjunto de datos de entrenamiento y validación. El modelo se utilizó para detectar noticias falsas sobre covid-19 en un conjunto de noticias recopiladas y clasificadas por los autores de este trabajo. La tasa de acierto fue del 70%, es decir, esta fue la tasa de éxito de detección de los artículos catalogados.
Assuntos
Humanos , Linguagens de Programação , Inteligência Artificial , Comunicação , COVID-19 , Desinformação , Coleta de Dados , Notícias , Troca de Informação em SaúdeRESUMO
Low power atmospheric pressure inductively coupled thermal plasma sources integrated with a quenching device (cold ICP) for the efficient production of biologically active agents have been recently developed for potential biomedical applications. In the present work, in vitro experiments aimed at assessing the decontamination potential of a cold ICP source were carried out on bacteria typically associated with chronic wounds and designed to represent a realistic wound environment; further in vitro experiments were performed to investigate the effects of plasma-irradiated physiological saline solution on eukaryotic cells viability. A thorough characterization of the plasma source and process, for what concerns ultraviolet (UV) radiation and nitric oxide production as well as the variation of pH and the generation of nitrates and nitrites in the treated liquid media, was carried out to garner fundamental insights that could help the interpretation of biological experiments. Direct plasma treatment of bacterial cells, performed at safe level of UV radiation, induces a relevant decontamination, both on agar plate and in physiological saline solution, after just 2 min of treatment. Furthermore, the indirect treatment of eukaryotic cells, carried out by covering them with physiological saline solution irradiated by plasma, in the same conditions selected for the direct treatment of bacterial cells does not show any noticeable adverse effect to their viability. Some considerations regarding the role of the UV radiation on the decontamination potential of bacterial cells and the viability of the eukaryotic ones will be presented. Moreover, the effects of pH variation, nitrate and nitrite concentrations of the plasma-irradiated physiological saline solution on the decontamination of bacterial suspension and on the viability of eukaryotic cells subjected to the indirect treatment will be discussed. The obtained results will be used to optimize the design of the ICP source for an effective production of reactive species, while keeping effluent temperature and UV radiation at values compatible with biomedical treatments.
