Abstract
The COVID-19 pandemic had major impact on all worldwide traveling in 2020 and 2021, which also affected Muslims' ability to perform the ritual of Hajj and Umrah. Due to the pandemic, the Saudi government limited number of pilgrims to ensure their safety and prevent the spread of disease. This reduction of pilgrims affected residents of Makkah and Madinah holy cities, whom were often involved in a variety of activities during the Hajj seasons. Twitter was one of the channels that many people and government organizations utilized to communicate with each other regarding general issues as well as other topics related to the Corona epidemic and health measures during the 2021 Hajj season (1442 AH). In this study, more than 22,000 tweets were collected from Makkah and Madinah during the Hajj season of 2021. The collected tweets were analyzed using convolutional neural network and long short-term memory network (CNN-LSTM) deep learning model. We extracted the basic features of tweets and classified them into three sentimental categories including positive, negative, and neutral. The findings indicated strange negative feelings towards the Corona pandemic differently within the tweets from the holy cities. The results also showed strong correlation value of 0.74 for the sentiment analysis rates of general tweets sent by Makkah and Madinah tweeters. Furthermore, considerable correlations have been observed between Makkah and Madinah tweeters' perceptions of several issues relevant to the Corona pandemic, revealing worthwhile remarking results.