NATO conducted one of its largest exercises over the past decade, Trident Juncture, from October 21st to November 6th 2015. The exercise included over 36 thousand troops from over 28 allies, 9 partner nations, and engaged 18 observations and 12 international and non-governmental organizations and aid agencies. As part of this exercise, the Center for Computational Analysis of Social and Organizational Systems (CASOS) at Carnegie Mellon was asked to assess, in partnership with the Data Mining and Machine Learning Lab at Arizona State University, the social media response to Trident Juncture. We collected data from Twitter and VK to provide daily updates and intelligence reports on the social media discussion surrounding Trident Juncture and NATO. We focus on three distinct collections of Twitter data: geo-tagged tweets, tweets by Ministries of Foreign Affairs (MFAs), and news media tweets. This is our summary report for the time period from October 1st through November 15th. The report includes as appendices our briefings made during the exercise.
[1]
Eric Gilbert,et al.
VADER: A Parsimonious Rule-Based Model for Sentiment Analysis of Social Media Text
,
2014,
ICWSM.
[2]
Wei Hu,et al.
Twitter spammer detection using data stream clustering
,
2014,
Inf. Sci..
[3]
Wei Lo,et al.
A Bayesian Graphical Model to Discover Latent Events from Twitter
,
2015,
ICWSM.
[4]
George A. Miller,et al.
WordNet: A Lexical Database for English
,
1995,
HLT.
[5]
Vern Paxson,et al.
@spam: the underground on 140 characters or less
,
2010,
CCS '10.
[6]
Michael I. Jordan,et al.
Latent Dirichlet Allocation
,
2001,
J. Mach. Learn. Res..