With increase in technology and use of smart phones there has been a rapid increase in real-time social media in the past 5 years. Interest in leveraging social data-streams to readily detect the emergence of social event in real-time has garnered significant attention, especially as it relates to business intelligence. To this end, researchers at the University of Louisiana at Lafayette have developed a novel temporal-evolution model that employs divergence scores, a series of thresholds, and unsupervised graph clustering methodologies to detect an event onset. Using historical social media data as a test-bed, the current model detects the onset of events 3 minutes (on-average) after the first pertinent social media posting. Regarding Twitter feeds, the current model out-performs the Temporal and Social Terms Evaluation (TSTE) standard in both sensitivity and noise-reduction and is comparable in speed of detection. Thus, the current model offers a preferred platform for decision informatics as it related to emerging event detection in social media.