Application of Event Stream Processing: Data Mining the 2017 Boston Marathon

Presented by

Michael Onwi and Devin Pearson, Kennesaw State University

About this talk

The ability to analyze high-velocity data in motion is increasingly of interest to organizations and civic leaders, enabling access to real time information for decision makers through event stream processing. This is imperative not only in sensor driven industries like health care and manufacturing, but also for large scale public events with issues related to public safety, transportation and overall experience for tourists and residents. The Center for Statistics and Analytical Research at Kennesaw State University used SAS® Event Stream Processing to analyze tweets from the 2017 Boston Marathon and determine topics, patterns and trends to provide insight for future large scale events. Attributes of interest from each tweet such as contents, timestamps and locations were identified and captured. The output was then filtered by US time zones. Two counter windows, global and EST (where the Marathon was held), were used to count the number of incoming tweets in a 10 second sliding window. The information was then displayed on a dashboard. The results were analyzed over the timeline of the marathon to correlate Twitter activity to the event itself. In addition, key topics were analyzed to determine propensity for co-presence, such as pain+Advil.

Related topics:

More from this channel

Upcoming talks (0)
On-demand talks (99)
Subscribers (16243)
In today's organizations, you need to get relevant data quickly to drive faster business decisions. With big data, sometimes that's easier said than done. SAS® Business Intelligence offers predictive insights with the ability to understand the past, monitor the present and predict outcomes, no matter the size or complexity of your data. In fact, SAS helps you deliver accurate, valuable information – from Hadoop or any other big data source. Plus, SAS offers an integrated, flexible presentation layer for the full breadth of SAS Analytics capabilities: data and text mining, statistics, predictive analytics, forecasting and optimization. One of the key components of SAS Business Intelligence, SAS Visual Analytics, offers self-service data discovery, enabling even nontechnical business users to explore billions of rows of data in seconds. With this tool, you can discover more opportunities and make more precise decisions, easily publishing reports to the Web and mobile devices.