ABSTRACT
More than two years after the start of the coronavirus disease (COVID-19) pandemic, the whole world continues to be impacted by this global crisis. Indonesians use the social media platform Twitter to share information and opinions about coronavirus disease (COVID-19) vaccination. This study was conducted to determine the views of Indonesians toward the government's COVID-19 vaccination program and to test the capability of several machine learning techniques to classify sentiments expressed on Twitter. The performance of four machine learning algorithms was tested: the Naïve Bayes, k-Nearest Neighbors (kNN), Decision Tree, and Support Vector Machine (SVM) algorithms. The findings show that the SVM algorithm exhibited the best performance in terms of accuracy (92%) compared to the Naïve Bayes, kNN, and Decision Tree algorithms. A grid search technique was also used to optimize performance based on parameter settings in the algorithm used. © 2022 IEEE.
ABSTRACT
Small and medium enterprises (SMEs) have been heavily affected by the COVID-19 pandemic, so many have switched from offline to online businesses to survive. Unfortunately, not all SMEs can use technology well, making it difficult to market their products online. The presence of consultants in social media marketing is expected to help the SME problems. This study aims to determine the level of acceptance of SMEs toward social media consulting, which is divided into technological, organizational, and environmental perspectives. This study also presented ananalysis of constraints and recommended solutions to improve the level of acceptance of social media consulting. The method used is descriptive quantitative by distributing surveys to 50 SMEs in Indonesia. The findings show that the acceptance of SMEs in Indonesia to social media is included in the highcategory, namely 80% of the total respondents, with the category of technology perspective at 88%, organization at 72%, and environment at 88%. © 2022 IEEE.