User Identity Sketching: Diving Into the OCEAN Profiling Model
25 Aug 2020
There are around three and a half billion smartphone owners today, or 44.85% of the world’s population.
As surprising as that may seem, it’s not the number of users we have to concentrate on, but how much we rely on mobile devices. Smartphones have become the natural extensions of our hands ‒ for example, an average US adult spends close to three hours a day using them.
That, of course, brings some consequences to the way we consume information, and the way we broadcast it, mostly about ourselves. The time we spend with our smartphones keeps increasing, and this is not just bad for our general health or productivity, it is also a growing threat to our privacy.
The ever-rising usage of smartphones is a condition that allows companies to identify our personality traits with higher accuracy and consistency.
Our location data, communication patterns, and interaction activity, as well as our multimedia consumption, generate troves of personal data that is in extremely high demand. It makes it possible for advertisers to engage in user profiling that can reveal an awful lot about us, without us inputting any such information directly.
StealthTalk may not be able to protect people from the side effects of smartphone consumption, but its team feels obliged to inform everyone about the number of things our everyday helpers can tell about us, its owners.
If you ever wondered how deep user profiling gets, and how data is analyzed to construct a detailed user portrait ‒ you’re just in the right place for an explanation.
The Consequences of Effective User Profiling
To understand user profiling, we first have to ask ourselves why it is even a thing.
For the lack of a better example, in 2019, Facebook generated close to 69.66 billion dollars in ad revenue that relies heavily on effective user profiling. Ads are Facebook’s bread and butter, and when so much money is on the line, beneficiaries look to explore the newer ways to make the ad business even more profitable than it already is.
It has already evolved to the point of being a science of its own, leveraging psychological studies and statistical analysis, and combining them to create an all-new field ripe for exploration.
Advertising strategies are becoming more sophisticated by incorporating artificial intelligence to recognize behavioral patterns and analyze collected digital footprints to create psycho-demographic user profiles. While that may sound very progressive and revolutionary, this is something that has been around for decades, just not at a bigger scale than today.
One of the models used often for this purpose is the OCEAN profiling model (also known as the Big Five model) created in 1980 to evaluate the influence of personality traits on academic performance.
The most recent use case of the OCEAN-associated model can be linked to Cambridge Analytica, a firm involved in "psychographic profiling" concerning the 2016 US presidential elections. Who knew that a simple quiz that offered Facebook users to find out more about their personality had anything to do with politics?
It did the trick and could do it once again. That’s why we are talking about it.
So What’s in the OCEAN of Emotions?
OCEAN model is a suggested taxonomy of five main personality traits:
- Openness (to imagination, experience, feelings, aesthetics, actions, ideas)
- Conscientiousness (competence, love of order, discipline, caution, ambition)
- Extraversion (assertiveness, dynamism, friendliness, adventurousness)
- Agreeableness (willingness to trust, modesty, genuineness, helpfulness)
- Neuroticism (self-control, prevalent mood, self-consciousness, emotional stability)
As the new research from Princeton University states, smartphones can unveil four out of five personality traits, based on communication activity, music choices, app usage, daytime, and nighttime activity levels.
For example, your phone’s average battery level is used to evaluate if you love order and have a sense of duty.
The average length and number of text messages, emoji usage, number of incoming and outgoing calls can reveal how sociable you are. Heightened nighttime activity and the number of device unlocks can indicate your low discipline levels. The average volume level when listening to music, your habits of swiping through the playlists, and frequent changes to it can underline a lack of caution, patience, or just a strive for newer things.
Things like weather checks can unveil your preparedness levels, and your habits as a photographer can indicate that you are an outgoing, open, and active person.
Smartphones still have a bit of a problem finding out how agreeable we are, at least not reliably, but those can be identified through additional factors. Imagine that you are on Youtube, and a content creator asks you to like the video before viewing it.
Would you agree to it? Yes? Then there’s another tick in your social profile.
Of course, the algorithms started off estimating basic traits that didn’t require advanced recognition methods to provide analysis. Then more and more unobvious observations were added in attempts to make the profiling procedure more precise.
But how accurate are those predictions?
There’s Too Much Plastic in the OCEAN
Understandably, a lot of people refer to this profiling activity as a pseudoscience that can not be taken too seriously.
“Big data is a dangerous, faith-based ideology. It's fuelled by hubris, it's ignorant of history, and it's trashing decades of progress in social justice.” ‒ Stilgherrian, freelance journalist, commentator, and podcaster.
With that, people are no strangers to carrying out critical decisions based on so-called metadata ‒ a set of data describing or giving information about the user or other data. Something that doesn’t provide the answer outright, but implies it by logic.
NSA General Counsel Stewart Baker has once said, “metadata absolutely tells you everything about somebody’s life. If you have enough metadata, you don’t really need content.” The same could be said about smartphone usage. There’s no need to ask any questions ‒ mobile devices will tell more about the person than they would themselves know.
No matter how much we may dislike it, the world is used to putting labels on people to categorize them.
Will this analysis ever be 100% accurate? It never will be, just like people. However, it would still give interested parties a good understanding of the user habits and preferences that could be taken into consideration when offering some product, or influenced for a bigger purpose, like elections. Here’s something to think about:
“The Internet exists for three things. It tries to sell you things, it tries to steal from you, and it is used to manipulate your opinion about different things. The more that you use the Internet, the higher the chance is that you will become a casualty.” ‒ Michael Hamilton, The Art of Email Security.
Based on this, we can get depressed and assume that what we are doing to decrease and minimize our digital footprint is mostly pointless. We can also choose to disregard the implications this model has for our privacy, citing that we are "just a drop in the ocean," but that's also not entirely true.
They say water is the softest thing, and it is also the most powerful, which makes it very similar to data. While there’s strong evidence that suggests the OCEAN model can be used for targeting to influence people's actions, whether it is purchasing or voting decisions, people could also call it “digital phrenology” and be right.
The absolute truth is somewhere in the middle.
One thing is for sure ‒ there’s too much we don’t know about OCEAN profiling to assume it is safe, and we have to understand how much information about us can be extracted through secondary indicators.
Hopefully, this blog post did raise some awareness for our readers. As some say, forewarned is forearmed. We would also like to forearm you not only with information, but something that could protect it ‒ StealthTalk. It is designed to make your phone calls and messages a private business, saving the most sacred personal information only to you.