“Smart gifting” could be the long-awaited answer for shoppers looking to buy the right presents for difficult recipients.

    With the help of deep learning, startup ebo box finds gifts that combine personal preferences with data from the general market, linking the traits that correspond to particular items or experiences.

    The neural networks were trained on a database of 21 million survey participants, and creators say this new technique can spot connections that humans tend to miss to find the perfect gift.

    According to the firm, ebo box is an “intelligent, interactive experience that helps pick a one-of-a-kind*, perfect gift for you.” It will allow users to designate a price they’d like to pay per gift, according to the Nvidia blog.

    To use the service, gift-givers must set up a profile and answer questions about themselves and the recipient. This includes basics like gender, age, and location, along with interest in activities, whether the gift-receiver is introverted* or extroverted*, and whether they are strictly logical, or tend to be more creative.

    With this information, the algorithms* may spot “hidden connections.”

    One example of this shows that people who drink a particular type of soda, prefer the color green, and love hiking, are also 52 percent more likely to buy a specific brand of smartphone.

    (SD-Agencies)