When Words Evolve, How Can You Understand a Page?

 

 

“If meanings of words can evolve due to conversations that are happening on a global scale, it follows that pure identification of words in an article (or keyword targeting) isn’t an effective way of truly understanding context.”

 

Collins Dictionary recently released their words of 2018 and they make for sombre reading. Not least in making me feel decidedly uncool – ‘flossing’ seems to have changed from something I lie to my dentist about to a dance that I’ve seen kids doing at family weddings – but also in shining a light on the biggest issues in 2018. They cover the gamut of environmental concerns around plastic with ‘single-use’, Brexit panic with ‘backstop’ and the growing cultural movement with ‘MeToo’.

Interestingly, the majority of the words featured on the list aren’t completely new to the English language. Collins has identified where there have been core changes in the meaning of the word due to ongoing, dynamic conversations. Take the word of the year, ‘single-use’: the term has gone from being purely descriptive of function or limitation to one that is loaded with environmental panic, bringing to mind David Attenborough and images of oceans overflowing with plastic.

How meanings and contexts change is something that illuma has been researching with our team of computer scientists. If meanings of words can evolve due to conversations that are happening on a global scale, it follows that pure identification of words in an article (or keyword targeting) isn’t an effective way of truly understanding context. And if that link is broken, how can we use it effectively to target new audiences?

photo-1513957723230-c330c6152342.jpg

The new context

In fact, illuma has found that a multidimensional approach is required to accurately understand a page online. Our system looks at a plethora of data points to score an average page’s contexts, not just a single category such as keywords, using advanced machine learning techniques to produce the most advanced content categoriser in the market.

If we look at another word on the Collins Dictionary list, our system can identify when ‘gammon’ is on a page online talking about what to have for your Boxing Day lunch, or when it’s being used in a somewhat inflammatory way to describe a pompous Brexiteer. And when it comes to advertising, a decision can be made accordingly on whether this page is the right place to run a brand’s campaign.

Word meanings shift as language and society naturally evolves. Illuma is developing systems using AI and machine learning to identify the true contexts as they change – data that radically improves online campaign performance.

Now, I’m off to do a bit of plogging (‘a recreational activity, originating in Sweden, that combines jogging with picking up litter’). See you next time.


illuma uncovers audiences and new prospects through mood targeting, without personal data – get in touch for more info, or to set up a product trial.