Who we are

Relative Insight’s history is a fascinating one, as we were born into the most demanding text analysis environment possible – law enforcement.

Initially, we were part of a collaboration between academia and UK law enforcement agencies to expose online criminal masquerading behaviour. Specifically, in a child protection context.

Children have a limited grasp of the concept of online personas and this makes them vulnerable. Our job was to deploy technology that could tell the difference between children and those people pretending to be children, and so we had to be able to know whether someone was really a 13-year old girl, or if they were a 40-year old man doing a sophisticated impression.

To do this, we deployed methods of comparative text analysis that we had been developing. It’s difficult to manually spot the difference between the language of a genuine child and the language of masquerading adult, so instead our text analytics platform would compare suspect language with known children’s language to isolate differences.

It was these differences that mattered. The small differences in words, phrases, grammar, and punctuation used by masquerading criminals are ones that could never be spotted manually.

This focus on doing comparisons and spotting differences continues to be a pillar of the Relative Insight philosophy.

By focusing on the differences between several sets of data, you discover insights that you would never have thought to search for – because after all, it’s all relative