Automating Military Intelligence Confidence Assessments for Twitter Messages

Some recent systems accurately produce high-level situational awareness by mining traffic in Twitter. Where these systems have been successful, there has been no attempt to evaluate Twitter streams for source reliability and information credibility because the situations have not been adversarial. The use of Twitter in recent political dissent in the Mideast makes the need for computationally tractable approaches to evaluating Twitter source reliability and information credibility more acute in order to produce accurate situation awareness in the face of misinformation or deliberate disinformation.