The TextQ system harnesses machine learning technology to process textual data in real time, in order to detect harmful (but rare) events and raise alerts. TextQ includes a feedback loop that ensures that, over time, fewer false positives will be identified. Thus, TextQ saves our customers time and money, and prevents psychological and personal harm from exposure to content that causes distress.

Other text filtering products send too many alerts which results in user fatigue and lack of faith in the system. By contrast, when TextQ sends the customer alerts about potential misconduct, it provides a mechanism for clients to give feedback about the usefulness of the alert. Over time, TextQ becomes personalized – alerting only about topics of interest to the user. In other words, customer interaction improves the performance of the TextQ system, resulting in better accuracy and less time wasted reviewing content that should not be flagged.

Parents and schools will benefit from TextQ by learning about harmful situations impacting children, so they can intervene appropriately and offer support. In the corporate environment, harassment and discrimination can be detected early, and corrected, resulting in fewer lawsuits and a better environment for employees and customers. Online platforms will reduce the time spent moderating content and reduce dependence on user reports of harmful posts.