403Stream Data Analytics for Multipurpose Social Media Applications
of speech obtained from a tagger developed by Carnegie Mellon University specically for Twitter
language usage instead of published text documents. The other is based on our patent-pending technol-
ogy SNOD. Our integrated algorithms provide much higher accuracy than the current approaches.
Figure 30.6 illustrates details of the threat prediction module (module 5 of Figure 30.2).
The pseudocode for the threat detection algorithm is as follows:
1. Identifying threatening tweets
Input: Set of tweets T, Bag of words A, B, C
Output: Set of threatening tweets X
For every tweet t in T,
If t contains a word from set B then let n = location of this word
If at least two words from ...