US 12,271,913 B1
Predictive social media alerting system
Ryan Austin Rasmussen, Chicago, IL (US); and Adam R. Khan, New York City, NY (US)
Assigned to Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed by Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed on Aug. 28, 2020, as Appl. No. 17/006,215.
Int. Cl. G06Q 10/0631 (2023.01); G06Q 30/0201 (2023.01); G06Q 30/0251 (2023.01); H04L 51/52 (2022.01)
CPC G06Q 30/0201 (2013.01) [G06Q 10/06312 (2013.01); G06Q 10/06315 (2013.01); G06Q 30/0252 (2013.01); H04L 51/52 (2022.05)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
receiving a set of social media messages referencing an entity during a time period;
comparing, using a processor, indicators from the set of social media messages to a baseline model to determine a likelihood of an upcoming event related to the entity to occur, the upcoming event having an impact on a resource of the entity, the resource including bandwidth available for accessing a website, wherein the baseline model is trained to output a prediction based on labeled data from previous events;
determining whether the bandwidth available for accessing the website is insufficient to provide access to the website based on the likelihood and the impact;
identifying, using the processor, a need for an increase in the resource of the entity based on the determination of the likelihood that the upcoming event is to occur after the time period, the determination indicating that the likelihood exceeds a threshold, and based on determining that the bandwidth available for accessing the website is insufficient;
predicting an end time for the upcoming event by comparing indicators from the set of social media messages to the baseline model, the indicators from the set of social media messages including a number of unique authors;
outputting an alert indicating the need for the increase in the resource of the entity and the predicted end time for the upcoming event;
automatically increasing, using the processor, the bandwidth available for accessing the website in response to outputting the alert; and
automatically decreasing, using the processor, the bandwidth available for accessing the website at the predicted end time.