Projects‎ > ‎

Social Media Analysis during Natural Disasters


Social media can be a great indicator of many global phenomenons in the world. However, analyzing the tremendous amount of content in order to extract useful information is still a challenging task.

We are currently developing on a parsing algorithm that will allow users to scan Social Media feeds such as twitter and decide whether the post is related to a Natural Disaster or is just a common expression.

Examples of this work can be found at the following location:


This website contains a sample analysis of  about half a million tweets related to violence, natural disasters, accidents, etc. Please note that this information is limited to geocoded tweets, which still represents a minor percentage of all tweets. Queries are limited to the top 1,000 tweets by keyword or by geographic area.