'Project Owl' - The Google Attack on Fake News And Wrong Content
‘Search can always be improved’ says Google as they release a new update for their search engine. In a recent blog post, they went through what their new update, titled ‘Project Owl’ and what it will be bringing in.
Throughout the years, Googles search engine has gone through many changes and updates. This was to combat websites that were trying to get better results through negative means such as using content farms and many other negative SEO techniques. Now Google have pushed their attention to mobile-first indexing and fake/wrong ranking content.
The most recent method people are trying to get around Googles ranking system is through fake news. This “news” can be misleading, low quality, offensive or can be false. This means that suggested searches (through Googles autocomplete feature when searching) and featured snippets are getting reviewed. Google are making quite a few changes to their search engine to tackle this, including general changes to the search rankings and adding more options for direct feedback from the public.
Overall, with this update, Google is launching:
- Feedback forms for search suggestions and formal policies about why suggestions might be removed.
- Feedback forms for ‘Featured Snippets’.
- Focusing more on authoritative content to improve search quality.
We will go into more depth on these points throughout this article.
Improving Search Quality
The idea behind Googles ranking algorithm is to help identify reliable sources of information which requires it to go through hundreds to billions of pages. This number is ever increasing due to more websites and pages being published every minute of every day. Every so often, the search algorithm goes through changes to combat unreliable or offensive content. Most searches will not return results that are reliable, but there are a small number of queries that still do.
Google stated that around 0.25% of search queries are still returning offensive or misleading content. To combat this, Google are making changes to the algorithm and its evaluation methods.
Search Quality Rater Guidelines Updated
Through using experimentation, Google have been making changes to the guidelines that determine a sites quality. Evaluators have been assessing Google’s search results give them feedback. From this feedback, Google can make more changes to the guidelines. When someone completes a search, they can flag a result they see as inappropriate. This item that is flagged could be anything from just false information to something much more offensive. The algorithm will then be able to get help from these guidelines to demote lower quality content. Furthermore, these guidelines will also aid Google in making changes in future.
Google have hundreds of different factors that can determine a sites rank. ‘From the freshness of the content, to the number of times your search queries appear on the page’. Google have adjusted these signals to ensure more authoritative content is shown. This will demote any content that is deemed to have low-quality content. There are obviously many other ranking factors that will determine a sites rank. However, currently Google are focused on inappropriate or fake content.
Direct Feedback on Autocomplete
Google detailed their aim when it comes to features like autocomplete. They wish to make sure that the experience is as quick as possible while also ensuring the suggestions are not offensive. Autocomplete is the function in which Google will offer suggestions of searches while you type. These are generated by an algorithm. It uses data through what other people have searched for and what is currently available online.
Using this feature however, can then bring up offensive or inaccurate information. Nothing could be done about this either. Now Google have changed that. They have added in a feature which enables you to report inappropriate predictions. Over time, this should remove inappropriate or offensive searches and content. These reports will then be used by Google to improve their algorithms.
Direct Feedback on Featured Snippets
Featured snippets are highlights of information that Google’s algorithm feels are relevant to what you are looking for. This brings up a small box at the top of the results with some information on your search. This snippet is pulled from the website that it originated from. Again, like the autocomplete feature, this has been bringing up offensive or inappropriate content. To combat this Google have implemented the same option as with autocomplete. There is now an option to report the content that is being shown.
The image below shows what it looks like when a search snippet appears. These can be with or without an image dependant on the search performed. The option to report it is then below the right side of the snippet. From here you can then report that snippet of information. Google will then be able to use this information they receive to improve the results.
Googles Improved Transparency
Within the blog post that Google published, they spoke about their transparency. This was in regards to their ranking system and other features. Through receiving many questions by the public, they have compiled everything to do with the new autocomplete policy into their Help Center. This was to give people more information about autocomplete and what Google are doing to improve the results.
Furthermore, Google published a ‘How Search Works’ site. This contains information on the technology behind their search algorithm. This is for people that would like to learn more on how Googles ranking systems sort through hundreds of billions of pages.
The Conclusion of ‘Project Owl’
This project is all about getting feedback from the users of Google. It means that the public will be able to openly aid in the improvement of the search engine. This will mean less inappropriate, offensive and false information will be found online. Furthermore, with Google being more transparent with what is behind their systems, it offers companies more understanding on what is going on in the background.Back