iconeWe’re here to bring you signal from noise with quantitative semiotics and machine learning.

quantitative semiotics machine learning platform how it works

Case study: the narrative tone of Brexit

We deployed Signoi to analyse of thousands of ‘reasons to vote the way I did’ in the 2016 UK EU Referendum. Signoi quickly identified 7 core narratives from the text alone. Group 3 emerged as strongly authoritarian and very anti-immigration. We called it ‘Get off my land’. Emotionally, it upweights on the language of fear and disgust. Archetypally, it is all about the Ruler: control. Unsurprisingly, this topic was dominant amongst the Leave voters.

brexit dashboard

Case study: Semantic Segmentation in Retail Banking

The Brexit work also can be seen as a form of Semantic Segmentation. This can be deployed commercially to good effect. Semantic Segmentation provides an escape from the old world of lengthy questionnaires, and lets you segment people faster, more effectively, and for way less money than traditional research methods.

Ask people to tell you in their own words what they need from your brand. For example, their top five needs and wants. Signoi will do the rest.

Here’s a snapshot of some work in progress, highlighting the rich opportunities to develop new ways of looking at banking customers according to the needs they express.


Case study: dominant codes in the Health and Beauty category

An analysis of thousands of images drawn from the health and beauty category. Signoi found 12 emerging dominant semiotic energies. Again, we could also map the image dimensions and consequent clusters onto the Jungian narrative archetype framework: for example, Hero and Caregiver



We then deployed the methodology onto more complex advertising imagery from the fragrance category. Comparative analysis allows the emerging territories of communication to be mapped, showing commonalities and differences. A few outputs are shown below… again, we show the Archetype Engine as one example of the universal frameworks that can be brought to bear. There are, of course, many more.



Case study: Automotive

We’re also working on a run of decoding car adverts. Here’s a sneaky peek at some of the raw underlying structures. The platform can easily identify codes of (examples clockwise from top right) Technology, Glamour, Family, Hyperrealism, and Adventure. This is in the exploratory phase at the moment. We’re feeding the machine more and more… and the more you feed it the better it gets!


If you found these simple examples interesting, rest assured you’ve only just scratched the surface of what Signoi can do!

Why not start a conversation?

Email Us