Valencia (EFE).- A team of researchers from the Polytechnic University of Valencia has created an artificial intelligence model that is capable of predicting the age and sex of a buyer in a physical store based on their eye and body movements and their position.
The work has been developed by the CBR (Consumer Behavior Research) group of the Immersive Neurotechnologies Laboratory (LabLeni) of the Humantech Institute of the UPV, within the framework of the European project Rhumbo, and has been published in the Journal of Virtual Reality.
The study of artificial intelligence
This group of researchers has developed a highly realistic virtual store in which the study participants, equipped with a series of devices such as eye-tracking cameras or motion sensors, could move freely exploring the store and making purchases.
“With these devices we were able to record in real time how they looked at the products, moved their hands and heads, moved around the store… In short, we analyzed the unconscious behavior of each customer in our store and found a series of common patterns based on of their age and sex”, points out Jaime Guixeres, coordinator of the study led by the CBR group of the LabLeni-Humantech of the UPV.
The study involved 57 people and they all performed three tasks in two different shopping situations: in one of them, they moved freely around the store; and in the other, they had to look for specific products.
“We collect information such as how long they spent in the store, their level of interaction with the products, and the number of products they looked at. And, later, we used all this data to develop a model based on artificial intelligence that allowed us to predict the age and gender of the participants”, adds Jaime Guixeres.
Differences between sexes and ages
In its conclusions, the study indicates that men look at more areas of the display and cover more space at a higher speed, while women take more time at each of their stops and are quicker to detect areas of interest.
Regarding age, young people move faster through the store and interact with more products, while adults are more selective in the way they look and make more stops per minute to detect relevant information in the store.
Virtual reality and personalized shopping
The study confirms the ability of virtual reality to collect information about people’s behavioral responses when shopping and how this information can be used to identify their demographic characteristics, in this case, age and gender.
“Having this knowledge gives brands the ability to analyze which areas of a hypothetical online store could be further personalized based on shopper demographics, to personalize products, enhance the shopping experience, and ultimately increase prices. sales”, Guixeres points out.
The CBR-LabLeni researcher at the Humantech Institute of the UPV highlights that this work represents a new approach in the study of consumer behavior that is more integrated and aligned with research on the metaverse in the retail or retail sector, tourism and marketing.
“It is the starting point for inferring not only gender and age, but also other more complex characteristics, such as our personality, in order to offer an increasingly personalized virtual experience”, concludes Jaime Guixeres.
The conclusions of this artificial intelligence study will be presented during the final meeting of the Rhumbo project, a neuroeconomics project funded by the European Union in which six countries have participated, which will be held on April 20 and 21 at the Polytechnic City of Innovation, Science Park of the Polytechnic University of Valencia.