Tweets posted during the Soma mine disaster that took place in Turkey provide a better understanding of the potential that social networks have for social support. The social media analytics framework and the text categorization methodology using CovNets were used to analyze 6.3 million tweets containing the keyword “soma” posted between 13 May 2014 and 23 March 2015 in Turkish. According to the findings, people used Twitter more after hearing about the terrible and tragic event. In reaction to the demand for a public day of mourning, Twitter was used to express grief and outrage. Twitter usage has expanded in unison with the involvement of charity and assistance organizations. Regardless, none of the support efforts posted on Twitter garnered widespread public participation. The results, on the other hand, showed that deep learning could accurately predict if a tweet would garner a substantial number of retweets. The findings could be beneficial for people interested in how social support organizations and policymakers use Twitter to keep the public informed during significant disasters.