Google has started using a new machine-learning Artificial Intelligence system called “RankBrain”. It is used to process search results in the Google search engine.
Machine learning is a situation where a computer teaches itself on different matters in place of being taught by humans.
RankBrain does not replace the latest search engine algorithm from Google which is Panda 4.2. It forms part of the overall search algorithm which is used to generate the most relevant links for the end user.
The use of RankBrain was confirmed by Google on 26th October 2015. In terms of importance, RankBrain is number 3 on the ranking algorithm checklist after links and content.
As an example, if RankBrain comes across a word or phrase for the first time, the machine will make an educated guess in terms of other words/phrases which might have a similar meaning and the results will be filtered accordingly. This makes it a perfect tool for scenarios where search engine users have typed in “never-before-seen” search queries.
Google’s total search algorithm has over 200 ranking factors which make up the ranking algorithm and the exact functions for each of them will never be fully disclosed by Google. The most widely accepted consensus is that RankBrain interprets the users search query to find matching pages that may not have contained the exact words typed in by the end user.
When offline, RankBrain is passed a batch of past search items and learns by itself when it matches this with the search results. Once the output from RankBrain is verified by an internal Google team, the system is updated and brought online again.
A performance and result orientated professional with extensive experience and proven success in building relationships.
