1st Runner-up in Edelman Predictive Intelligence Competition (Epic) Data Challenge 2019

About the competition: With Natural Language Processing & data analytics methods in the communication industry, we experimented with news & social media data and detected digital news patterns.

Date of the competition: 6 November 2019 

Achievement: 1st Runner-up 

Prize:  $1000

REP Students: Samuel Ang, Benjamin Chew, Aaron Sng

Product/Proposal: The analysis involved using Latent Dirichlet Allocation and Bag of Words method to determine topics and recurring issues. Their recommendations involved using the Stifler, Infected and Recovery (SIR) Model to generate Digital Twin models for prediction of future PR disasters.

Description: When a digital crisis occurs, every second counts. Brands must act swiftly - from drafting an official statement to managing the grievances expressed by their customers and wider community.  

 

For example: A passenger refused to give up his seat on the over-booked United Airlines Flight 3411. He was then forcibly removed from the plane, receiving severe injuries in the process. The video was widely circulated on social media, and the incident was reported in major news outlets – the story went viral and led to a huge PR crisis for United Airlines.  

 

In a hypothetical scenario, it is critical for both United and Edelman to understand the digital crisis landscape - the kinds of content that are shared, the individuals and publications doing the sharing, and how this process of information spread occurs. These insights will go on to drive informed crisis response and subsequent reputation management strategies.