ABSTRACT
We present here the results of a five-years-long earthquake educational project aiming to commemorate the hundredth anniversaries of five large Northern Apennines earthquakes occurred between 1916 and 1920 in the areas of: Rimini (1916, Mw 6.1), Valtiberina (1917, Mw 5.9), Romagna Apennines (1918, Mw 5.9), Mugello (1919, Mw 6.3) and Garfagnana (1920, Mw 6.5) earthquakes. We saw these anniversaries as the occasion for leading the Northern Apennines people to rediscover their past, in a positive way, and to improve their awareness of the earthquake as a natural feature of the regions in which they live. The activities that we planned for schools students encouraged them to go hunting for traces of the earthquakes of one hundred years ago in their home towns and to rediscover the memories and traditions of their communities. Together with their teachers, we also led the teenagers to find creative ways to involve the grownups in the process of discovery and knowledge. The Project had to cope with two emergencies: the great Central Italy earthquake of 2016-2017 and the Covid-19 epidemic. However, these stumbling blocks did not deter teachers and students from taking part in the process actively and even enthusiastically. Their families and communities were actively involved too. This experience taught us some valuable lessons. First of all, we learned to adapt the project, as we had conceived it at the start, to a wide gamma of social and cultural contexts. Not all the involved communities were equally aware of the level of seismic risk they are exposed to. On the affluent Adriatic coast, where tourism is the main source of income, past earthquakes are something best forgotten, by citizens and administrators alike. On the contrary, in the poorer inland mountain areas (Forlivese Apennines, Mugello, Garfagnana and Lunigiana) a more down-to-earth attitude prevails: earthquakes are looked upon as something that can and does happen and people are quick to grasp how important it is to contribute to initiatives whose aim is reducing seismic risk. Thus, we had to adapt our approach to the different contexts, modifying each time the activities we proposed to the schools with the aim of obtain the best possible results from each situation. © 2022 Istituto Nazionale di Geofisica e Vulcanologia. All rights reserved.
ABSTRACT
We present KAAPA (Knowledge Aware Answers from PDF Analysis), an integrated solution for machine reading comprehension over both text and tables extracted from PDFs. KAAPA enables interactive question refinement using facets generated from an automatically induced Knowledge Graph. In addition, it provides a concise summary of the supporting evidence for the provided answers by aggregating information across multiple sources. KAAPA can be applied consistently to any collection of documents in English with zero domain adaptation effort. We showcase the use of KAAPA for QA on scientific literature using the COVID-19 Open Research Dataset.
ABSTRACT
Neural passage retrieval is a new and promising approach in open retrieval question answering. In this work, we stress-test the Dense Passage Retriever (DPR) - -a state-of-the-art (SOTA) open domain neural retrieval model - -on closed and specialized target domains such as COVID-19, and find that it lags behind standard BM25 in this important real-world setting. To make DPR more robust under domain shift, we explore its fine-tuning with synthetic training examples, which we generate from unlabeled target domain text using a text-to-text generator. In our experiments, this noisy but fully automated target domain supervision gives DPR a sizable advantage over BM25 in out-of-domain settings, making it a more viable model in practice. Finally, an ensemble of BM25 and our improved DPR model yields the best results, further pushing the SOTA for open retrieval QA on multiple out-of-domain test sets. © 2021 ACM.
ABSTRACT
OBJECTIVES: The management of healthcare workers (HCWs) exposed to confirmed cases of coronavirus disease 2019 (COVID-19) is still a matter of debate. We aimed to assess in this group the attack rate of asymptomatic carriers and the symptoms most frequently associated with infection. METHODS: Occupational and clinical characteristics of HCWs who underwent nasopharyngeal swab testing for the detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in a university hospital from 24 February 2020 to 31 March 2020 were collected. For those who tested positive and for those who tested positive but who were asymptomatic, we checked the laboratory and clinical data as of 22 May to calculate the time necessary for HCWs to then test negative and to verify whether symptoms developed thereafter. Frequencies of positive tests were compared according to selected variables using multivariable logistic regression models. RESULTS: There were 139 positive tests (8.8%) among 1573 HCWs (95% confidence interval, 7.5-10.3), with a marked difference between symptomatic (122/503, 24.2%) and asymptomatic (17/1070, 1.6%) workers (p < 0.001). Physicians were the group with the highest frequency of positive tests (61/582, 10.5%), whereas clerical workers and technicians had the lowest frequency (5/137, 3.6%). The likelihood of testing positive for COVID-19 increased with the number of reported symptoms; the strongest predictors of test positivity were taste and smell alterations (odds ratio = 76.9) and fever (odds ratio = 9.12). The median time from first positive test to a negative test was 27 days (95% confidence interval, 24-30). CONCLUSIONS: HCWs can be infected with SARS-CoV-2 without displaying any symptoms. Among symptomatic HCWs, the key symptoms to guide diagnosis are taste and smell alterations and fever. A median of almost 4 weeks is necessary before nasopharyngeal swab test results are negative.