The definition of an adjective is “a word naming an attribute of a noun such as sweet, red or technical” .
So, it’s no surprise that a machine learning analysis of 3.5 million books found adjectives for women were physical and for men, their adjectives referred to behaviour.
“Beautiful” and “sexy” are two of the adjectives most frequently used to describe women.
On the other hand, commonly used descriptions for men include righteous, rational, and courageous, all non-physical attributes.
This is important, because this computer analysis will add to the computer knowledge behind machine learning or ‘AI’.
And these machines will impact our lives in the years ahead.
As part of their analysis, the researchers extracted adjectives and verbs associated with gender-specific nouns (eg “daughter” and “stewardess”).
For example, combinations such as “sexy stewardess” or “girls gossiping.” They then analysed whether the word combinations had a positive, negative, or neutral sentiment, and then categorized the words into semantic categories such as “behaviour,” “body,” “feeling,” and “mind.”
They found that verbs associated with body and appearance appear five times as often for female figures as male ones.
And here is the thing.
Many of the books analysed were published several decades ago, so the attitudes and stereotypes of say the 1980s are being learnt by the ‘machines.’
This is the technology that allows smartphones to recognize our voices and enables Google to provide keyword suggestions.
The algorithms work to identify patterns, and whenever one is observed, it is perceived that something is ‘true’. If any of these patterns refer to biased language, the result will also be biased.
As AI and language technology become more prominent across society, it is important to be aware of gendered language.
Interestingly, the machine used in this research couldn’t distinguish between genres—eg between romance novels and non-fiction.
Maybe when it comes to romance all adjectives are acceptable.
