Trait Prediction Engine
- Facebook LikeIDs
- Facebook statuses
- Browsing data
- Open text
- And more…
- BIG5 Personality
- Life Satisfaction
- Political Views
- Religious Views
- Relationship status
Our Vision Is To Personalise The Internet
Ever bought something from Amazon and seen it advertised in your Facebook newsfeed a week later? Or returned from a holiday you booked in person to find the identical flight offers in your Google search results?
Your online experiences are constantly tweaked and tailored based on an algorithm’s best guess at what you might want in a given moment. Most data-driven marketing tools try to build up a picture of you by collecting as much of your past behaviour as possible – scraping, buying and occasionally asking for whatever information they can get.
Apply Magic Sauce is different. We focus on the psychological traits and emotions that drive a given behaviour, not merely the fact you have clicked or Liked something. We care not about identifying you personally, but providing you with tools to individualise your own experience.
We encourage all of our collaborators to adhere to the following ethical principles, in addition to the applicable legal restrictions:
- Control: Nobody should have predictions made about them without their prior informed consent
- Transparency: The results of any predictions should be shared with individuals in a clear and understandable format
- Benefit: Predictions should be used to improve services and provide a clear benefit to users
- Relevance: It should be clear why the data requested is relevant to the prediction being made
How It Could Look
Apply Magic Sauce API enables conversation between businesses and consumers about how predictive technologies ought to be used. This could help algorithms explain themselves and learn from their mistakes, just like humans.
Choose a character to discover how their personality might affect their online experience, or click Predict My Profile to see how others see you.
How it works
Send us any digital footprint
And we will return a psycho-demographic profile
Our models are based on over 6 million social media profiles and matching scores on psychometric tests. This gives us the unique ability to prove how our predictions compare to validated tools that have been in use for decades. We publish our results in scientific journals and open-source our research data. We hope this approach will help stimulate an open debate about the future of predictive technology and encourage more user-driven service innovation.
We put people before predictions
Making artificial intelligence more attentive
Machine-learning algorithms can be difficult to interpret, let alone implement in a product. Without psychological sensitivity, these algorithms are at risk of perpetuating historical prejudice, being overly domain-specific or prescribing norms that can feel impersonal. We know that patterns of human activity are usually not random, so we always think beyond the mere clicks or Likes of an individual to consider the subtle attributes that really drive their behaviour.
Add psychological data points to any sample
Trait predictions turn gaps in your knowledge into actionable insights
Connected devices are recording everything we do and want - our pulse, our purchases, our personality, it’s all part of the internet. The websites we visit and the brands we interact with can certainly be very informative, but these are noisy snippets of behaviour, not the whole picture. Our psychometric techniques can identify and fill the gaps in your knowledge about yourself or your customers. We pave the way for collaborative tools through which citizens can better extract informational and commercial value from their own data.
Developed at the University of Cambridge Psychometrics Centre
Apply Magic Sauce is the product of a rich dataset, a young multi-disciplinary team and a supportive academic environment. It was assembled by researchers at the University of Cambridge Psychometrics Centre and builds upon a 30-year legacy of leadership in advanced psychological measurement and computational behavioural science.