One of the big questions search engines and internet users face is how do we determine how factually accurate and reliable a website is? Look for any fact on the internet and the chances are that you will find a wide variety of claims and information presented as the truth. You may need to check several sources to establish what’s correct.
On 12th February 2015, eight Google engineers published an academic paper entitled: Knowledge-Based Trust: Estimating the trustworthiness of web sources. In this paper, the engineers outlined their extensive work to create an algorithm that can accurately estimate how reliable a website is based on the factual information quoted on that website.
In the past, search engines have looked to external signals, such as hyperlinks to source material and browser history, to establish trustworthiness. This method has its flaws. Gossip websites, for example, may link to external sources and attract a great deal of traffic but how reliable are the stories they publish? Political groups can post propaganda as facts and the onus is on the reader to disprove those facts by cross-referencing with other sources.
On the flipside, there are many factually accurate, informative websites that struggle to attract traffic and may not currently receive recognition from the search engines.
How Knowledge-Based Trust would work
With Knowledge-Based Trust (KBT), Google proposes to crawl trusted websites to create a standard of accurate information, i.e. a database of proven facts. The KBT algorithm would then compare the claimed ‘facts’ on an individual website with the established facts in the knowledge base; this would generate a KBT score. Google would reward high-KBT sites with higher rankings and penalise low-KBT websites with lower rankings. In turn, high-KBT websites should receive more web traffic, improving many of the other signals that Google looks at, such as bounce rate and time spent on site.
Knowledge-Based Trust so far
The academic paper published by Google’s engineers shows that they crawled millions of web pages and extracted 2.8 billion trusted facts. They then compared these facts to data they found on 119 million web pages using the algorithm to establish KBT scores, then manually checked a subset of the results to see whether the algorithm worked. This manual check found that the KBT scores were a highly effective way to rank trustworthiness.
The paper also suggests that Google could create a KBT score for domain trustworthiness, as well as at a page level. This might show, for example, that certain news sources are more trustworthy than others, making them a more reliable source to turn to for news and fact checking.
The authors of the Knowledge-Based Trust paper say that KBT would provide ‘an additional signal for evaluating the quality of a website’ and could be used ‘in conjunction with existing signals such as PageRank’ rather than as a replacement. This means that websites without lots of factual information would not be penalised with a low-KBT.
A new era of trust
It makes sense that search engines are examining ways to check and reward trustworthiness. So much misinformation is spread online and Google is seeking to reward websites that have checked their facts. This is particularly important for niches were a few sites control most of the public information or for stories about medical research, for example. It may also facilitate a move away from link metrics as a ranking factor.
Trust and your website
Although Knowledge-Based Trust is an exciting possibility, there are certainly some issues that Google will need to iron out before the algorithm becomes a reality. Challenges such as filtering out ‘irrelevant noise’ and trivial facts, or developing sufficient extraction technology have all been raised as potential obstacles to Knowledge-Based Trust.
Even if Knowledge-Based Trust is months or even years away from being rolled out, it’s a timely reminder that every business should be focused on demonstrating its trustworthiness. If you present factual information on your website, it’s important to check and double-check that this information is correct, quote your sources, and back up your claims. Get caught giving out misinformation and it could cast a shadow over your whole business.
If your customers know that they can rely on your website for the facts, it will build your reputation. People will come to you for information, quote you in their own articles, and view your business as a trustworthy one that takes time to substantiate its information; they’re also likely to spend more time on your website, link back to it in blogs, and go deeper into your pages. This can only be good for your business – and the fact that your website’s trustworthiness might boost its SEO as a ranking signal one day is a happy bonus.
How do you fact check information on your website? Do you quote your sources and link back to them to help people read around a topic? I’d love to hear your thoughts in the Comments below.
If you enjoyed this post, we’d love it if you shared it – thank you