As social media has progressed, users have increasingly posted during disasters and produced infrastructure damage data. Gathering social media data has its limitations, but leveraging this data can prove incredibly enlightening and useful, due to the wealth of information. This study focuses on Twitter data related to Hurricane Florence, specifically classifying tweet images showing relevant infrastructure damage and identifying local journalist reporting on the hurricane. After some pre-processing, the image data was used to train a machine learning algorithm to classify the presence of damage, the type of damage, and the infrastructure damaged, among other labels. Additionally, Twitter user data was used to identify local journalists tweeting during the hurricane by assessing follower count and user bios. The goals are to use the machine learning algorithm to identify damage from social media images and to engage local journalists to contribute more information in future disasters.