ABSTRACT
Hope endows the power required to conquer life's challenges. Since the COVID-19 pandemic, hope speech detection has gained momentum in the natural language processing field due to the need for positive reinforcement on the internet. Hope speech detection identifies texts in social media comments that evoke positive feelings in people. This paper aims to identify hope speech in YouTube comments. To discern the hope speech as distinct from the YouTube comments, various machine learning and deep learning algorithms (Support Vector Machines, Logistic Regression, Convolutional Neural Networks + Bidirectional long-short term memory (Bi-LSTM), Multinomial Naive Bayes (MNB), Ensemble models) have been used. These YouTube comments have been created as a part of the task "EACL-2021: Hope Speech Detection for Equality, Diversity, and Inclusion". Based on the best configurations that showed the highest results in trials for the shared tasks, weighted F1-scores of 0.608 for Tamil and 0.911 for English were achieved. © 2022 IEEE.
ABSTRACT
The novel corona virus disease (COVID-19) has unfolded into a pandemic and is continuing to propagate at a frightening speed. The aim of this article is to share our protocol for performing a safe surgical tracheostomy in this COVID-19 era. Tracheostomy procedures have a high risk of aerosol generation. To standardize institutional safety measures with tracheostomy, we advocate using a dedicated tracheostomy protocol applicable to all patients including those suspected of having COVID-19. We also did explore the current literature and recommendations for tracheostomy in patients with COVID-19 and studied the previous data from severe acute respiratory syndrome coronavirus 1 (SARS-CoV-1), the virus responsible for the SARS outbreak of 2003. We have prepared a protocol for performing a safe surgical tracheotomy in patients affected by COVID-19. Surgeons who might be involved in performing the tracheostomies should become familiar with these guidelines.