Perhaps someone is blowing off steam in making a threat. By analyzing information over a period of time, officials can narrow the 100,000 posts that may be classified as critical to the 5,000 that could pose real threats and need further analysis. They can assess whether sentiment is changing over a period of time, either reducing or increasing the concerns of law enforcement about a serious threat.
When activity reaches a certain overtly threatening level, such as the desire to "hunt down a judge," then the system can generate warnings to law enforcement analysts such as Secret Service agents in the case of federal officials or private security agencies in the case of executives, entertainers, or actors.
Such notifications would be sent to requisite organizations for further evaluation. Another advantage of such a system is that technology can monitor events without human interaction, but we have to remember that no action can take place without a thorough investigation.
If a threshold is reached, the responsible agency must determine the validity and seriousness of threats before any action can be taken. This could prevent something more sinister from occurring. A person could be suffering from mental illness, and law enforcement can secure help before the person crosses a criminal line. Does the person have a violent background? The official may be able to diffuse the situation by simply talking to the person, and letting him or her know that someone is actually listening.
The goal should never be to stifle free speech. Instead, it should be to disrupt violent actions before they occur. Social media analytics technology is available and it's becoming more than just a way to communicate—it's a way to communicate with informants. It's now up to law enforcement and first responders to make the most of it.