Abstract

With over 16 million tweets per hour, 600 new blog posts per minute, and 400 million active users on Facebook, businesses have begun searching for ways to turn real-time consumer-based posts into actionable intelligence. The goal is to extract information from this noisy, unstructured data and use it for trend analysis and prediction. Current practices support the idea that visual analytics (VA) can help enable the effective analysis of such data. However, empirical evidence demonstrating the effectiveness of a VA solution is still lacking. A proposed VA toolkit extracts data from Bitly and Twitter to predict movie revenue and ratings. Results from the 2013 VAST Box Office Challenge demonstrate the benefit of an interactive environment for predictive analysis, compared to a purely statistical modeling approach. The VA approach used by the toolkit is generalizable to other domains involving social media data, such as sales forecasting and advertisement analysis.

Original languageEnglish (US)
Article number6820691
Pages (from-to)58-69
Number of pages12
JournalIEEE Computer Graphics and Applications
Volume34
Issue number5
DOIs
StatePublished - Sep 1 2014

Keywords

  • Bitly
  • Computer graphics
  • Graphics
  • Linear regression
  • Movie revenue
  • Movie reviews
  • Prediction
  • Social media
  • Twitter
  • VAST Box Office Challenge
  • Visual analytics
  • Visualization

ASJC Scopus subject areas

  • Software
  • Computer Graphics and Computer-Aided Design

Fingerprint Dive into the research topics of 'Business intelligence from social media: A study from the VAST box office challenge'. Together they form a unique fingerprint.

  • Cite this