The tool uses machine learning algorithms to analyse patterns in data and group them into common themes, but does not use code-frame based textual analysis.

Punctuation and peripheral words are removed by Maru/Hub Topic Modelling, which the company said leads to more "accurate" results.

Maru has worked with behavioural scientists at the University of Strathclyde to test the tool prior to launch.

Steve Brockway, chief research officer at Maru/Matchbox UK and lead researcher on the project, said: “Our next step is to take machine-based learning and apply it to analysis approaches to better understand what’s driving certain behaviours instantaneously.”