COVID-19 Risk Assessment of Occupations Using Interval Type 2 Fuzzy Z-AHP & TOPsIs Methodology

The pandemic spread of COVID-19 caused by a virus that affects the respiratory system represents a dramatic threat to life. People have been practicing social distancing by working from home in recent months since it is an excellent solution to reduce one’s exposure to COVID-19. However, many occupations do not have this chance due to the necessity to attend the workplace. Besides, some professions require close contact with infected people such as medicine and nursing, while others such as logging and gardening have a low level of risk in this respect. However, while occupations such as medicine and nursing can take precautions against the virus at a very good level, the knowledge of occupations such as logging and gardening is weak against it. Prioritizing occupations based on these conflicting criteria is an important task under the vagueness and impreciseness of human evaluations. In this paper, a novel fuzzy AHP & TOPSIS methodology is proposed, where the reliability and restriction of the expert assessments are considered by interval-valued type 2 fuzzy numbers. A real case study including five experts, 14 occupations, and seven criteria is presented. A comparative analysis is also given to validate the proposed methodology. © 2022 Old City Publishing, Inc.

On the Generation of Medical Dialogs for COVID-19

Under the pandemic of COVID-19, people experiencing COVID19-related symptoms have a pressing need to consult doctors. Because of the shortage of medical professionals, many people cannot receive online consultations timely. To address this problem, we aim to develop a medical dialog system that can provide COVID19-related consultations. We collected two dialog datasets - CovidDialog - (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19. While the largest of their kind, these two datasets are still relatively small compared with generaldomain dialog datasets. Training complex dialog generation models on small datasets bears high risk of overfitting. To alleviate overfitting, we develop a multi-task learning approach, which regularizes the data-deficient dialog generation task with a masked token prediction task. Experiments on the CovidDialog datasets demonstrate the effectiveness of our approach. We perform both human evaluation and automatic evaluation of dialogs generated by our method. Results show that the generated responses are promising in being doctorlike, relevant to conversation history, clinically informative and correct. © 2021 Association for Computational Linguistics.