Google is proposing to rank websites on factual information rather than how popular they are. How could this affect your site and the sites you search for?
The internet is full of information. Most of it is useful and can be found on factual databases, online libraries and scholarly journals. However, there is also a large portion of the internet that is made up of gossip and personal opinions. When it comes to Google, what gets displayed first on the search engine results?
Google has traditionally displayed results based primarily on popularity of terms. This is done by using hyperlinks to “crawl” between pages, eventually creating an index that can display which page is most relevant to your search. Google uses complex algorithms to generate these hundreds of thousands of results, displaying your “best” results first.
However, this method may change in the near future. Google research scientists have been working on a new method to use credible facts rather than just frequency of terms to display search results. This team of Google researchers has published a paper to explain their new idea of Knowledge-Based Trust (KBT).
The idea behind KBT is to display factual information before controversial opinions and lies, meaning no more gossipy information about celebrities popping up on page one of Google results. Instead, real, authenticated facts will be displayed at the top.
How these facts are confirmed will be done through a process referred to as “triples”. This is done by creating a pattern which is made up of three factors:
- a subject that’s a real-world entity
- a predicate that describes some attribute of that entity, and
- an object that is the value of the attribute.
For example, Jim Bezos is a subject; he is a CEO, which is the predicate; and he’s the CEO of Amazon, which is the object.
From there, the KBT algorithm uses a complicated multi-layer approach to determine which facts are actually true. Rather than just counting how many incoming links a site has, Google will now count the number of incorrect facts within a page, leading to a KBT score.
This new addition to Google’s algorithm will reference back to Google’s Knowledge Graph, sometimes known as the Knowledge Vault. This is described by Google as a database of 2.8 billion facts extracted from the web. This is nothing new: Google has been building this database since 2012. The Google Knowledge Graph has estimated the trustworthiness of 119 million web pages, with millions more expected to be added.
Google’s new method sparks fresh concerns for many. For example, what will happen to sites that aren’t entirely built on facts? There are millions of blogs that are based on people’s thoughts and opinions. There are entire online magazines that are based on gossip. Are these sites going to fall off the map?
Another concern is that many facts are actually just political statements. Will Google show an allegiance to one political party over another?
If Google does implement KBT into search results, it will not replace the existing method of displaying results. Instead, it will add to Google’s results and replace misleading information. It’s hard to say exactly how some of these sites will be ranked, but Google has made it known that this would be an addition, not a replacement.
Other sites such as online stores or portfolios contain no real factual information. These types of sites shouldn’t be affected by KBT and will continue to be displayed based on Google’s traditional algorithms.
It is important to note that this is a work in progress. As of now, Google researchers are determining exactly how it will work. Google has not released to what extent this will change classic search results. However, it is clear that Google will go to great lengths to start ranking sites by accuracy rather than just popularity.
Final thought? KBT is good news for research papers, but bad news for finding the hottest gossip.