Breaking Ground at Yahoo! Search

Search engines have come a long way since the days of old school Google and AOL, now we’re seeing search trends leaning in the direction of semantic search. Rather than looking at the keywords you’re using, search engines are learning to understand what you mean when you use search terms, and what it is you’re hoping to find with those particular keywords.

You can see an example of this with Google’s auto correct, while it used to just offer spelling suggestions, it can now determine what you mean when you mis-spell words. Search for ‘icland is an icland’, and you’ll see the following suggestion.

Scientists at Yahoo! have make ground breaking advancements into this field, bringing the possibility of truly semantic search even closer. They set out to provide answers, not links, as when we search the web we’re usually doing it to find an answer or solution to a question.

Rather than pulling keyword relevant links every time, Yahoo! search scientists have built a new search engine from the ground up which is able to learn your interests and preferences. Over time the search engine will understand what kind of links you want to find when you search and will know what interests you over what you ignore.

So far, this system is only functioning within searches for breaking news and entertainment as these are two areas which fare incredibly well when subject to individual preferences. If a user were to submit the same search query throughout the day, Yahoo! search can update the results based on advancements in the story, or changing relevance, based on user behaviour.

The question is of course, do we want this kind of personalisation, as this relies on a huge amount of loyalty to one search engine, and it also raises privacy issues. And what would this mean for SEO? Surely this would be another way to filter out keyword stuffing sites and promote valuable content creators.

Share your thoughts below!