REIsearch NEWS


Artificial intelligence is making it easier to describe smells

Most people don’t have the right words to describe what they’re smelling. Though humans can distinguish about a trillion odors, our vocabulary is limited. Terms  like “fruity” or “musky” are not only imprecise, but also colored by cultural bias. Unlike with other senses—hearing, sight, touch, and taste—we have trouble agreeing on universal terms for smells.

Now, an IBM study recently published in Nature suggests a promising solution to augment our smell vocabulary. Researchers led by computational neuroscientist Guillermo Cecci used artificial intelligence to create an algorithm that translates fuzzy descriptive words to their molecular equivalent, and vice versa.

Quartz reports how researchers summarized their findings in an “odor wheel” that takes the most commonly used English words to describe scent, and arranges them in associative order. For instance, “vanilla” is adjacent to “chocolate” and “caramel” on the wheel, signaling to perfumers and chemists that when someone wants a vanilla candle, they likely won’t mind hints of caramel or chocolate.

The odor wheel also shows that some descriptive words translate to molecules better than others. Words written in red, including “sweaty” or “cadaverous,” are understandable to chemists, while those in written blue, like “soapy” or “cardboard,” are problematically vague.

Twitter Feed