Social Sentinel’s proprietary way to scan (not monitor) public social media for harmful or threatening behavior combs through billions posts. That’s the easy part of what we do.
Many authors on public social media don’t readily volunteer their affiliation with their schools and don’t share their geographic location. Therefore, one of the challenging data science gaps we needed to solve was associating threatening posts with you in a meaningful way—one that contributes positively to your district or campus safety program.
Necessity is the mother of all invention.
Our solution to the challenge is a series of proprietary algorithms that increase the relevancy of our associations. Some of the critical inputs to these algorithms are characteristics unique to a client and their community; like the formal and informal terminology used to describe its locations, people, events, and more. Managing these inputs is done through a feature called People, Places and Events (previously known as Local+).
This feature provides an area within the Social Sentinel application to enter keywords into categories. The added keywords directly impact our system’s ability to create stronger associations for your school and community. Categories include:
Names of high-profile individuals, VIPs, visiting dignitaries, etc. (with their permission)
Location names, object names, colloquial terms for physical assets and areas related to a client
Time-based activities such as sports games, graduation, etc.