Abstract

During a crisis, people who are in an emergency can not report their need to the 911 system, because of the huge number of calls, so people tend to use social media to ask for help. The 911 system until this time doesn't have a tool to detect posts from social media and act upon it. This research will take part in a project that will make social media data available for the 911 system during crises. So, workers in the 911 system could detect posts related to a crisis and send the information to dispatchers needed to approach the people who are in danger. This research will mainly focus on the information related to shooting events and identify the most frequent keywords used by people during these events. The methods that are going to be used will be divided into three sections, collection, storage, and analysis of the data coming from the twitter stream. And then compare the results of the analysis of the twitter stream to the results of the analysis of existing shooting datasets that were collected during previous crises. Then, both results will be visualized using Python Programming language and identify the statistics of keywords that appear in tweets that contain actionable information.