Page 1 of 1

Google BERT Natural Language Processing Update

Posted: Thu Dec 12, 2024 7:36 am
by mstlucky7800
BERT (Bidirectional Encoder Representation from Transformers) is the largest algorithm update in 5 years, which covered about 1/10 of all searches in the world. Its main goal was to provide better and more relevant search results for long-tail phrases. By incorporating Natural Language Processing into the algorithm, Google gained the ability to better interpret content, and thus provide the search engine user with better tailored answers.

Since December 2019, BERT has been working for 70 languages ​​(including Polish), mostly for long-tail phrases, and for now we don't have to worry about any sudden changes for phrases with a clear searcher intent.

As Google itself points out, it is not possible to optimize pages for BERT. Decreases or increases are not a punishment or reward, but a result of the algorithm changing its understanding of user intent.

The introduction of BERT is a milestone on the way to receiving more relevant answers. More and more people enter longer phrases into the search field and expect a precise answer in the search engine (or at least a clear declaration that they will get such an answer on a given page). Moreover, in the era of rapid development of voice searches, algorithms must daily deal with formulations that resemble communication between people more than between humans and machines. The correct interpretation of the query determines whether the user will receive an answer to their question or will be left with nothing. Moreover, it is estimated that about 15% of all queries are completely new queries, which the algorithm must deal with relatively quickly in order to maintain a high level of satisfaction for Google's flagship service.

Launch date: October 2019 (English queries) December 2019 (expansion to 70 languages)
Goal: Better understand user intent and better tailor search results for long-tail phrases
Who Gains: Websites That Are More Responsive to Long-Term User Queries
Who lost: Sites whose content does not correlate with the user's intent.
January 2020 Core Update
Google started the new year with a bang. Already in the first half of January, information about the release of the core algorithm update appeared on Twitter. The giant from Mountain View is continuing the tradition started in March 2019 and giving subsequent updates their own names. This one was named January 2020 Core Update.

From the brief information posted on the official Google Twitter account, we can only learn that the instructions on how to proceed in the event of an algorithm update remain unchanged.

The January update of the main algorithm - like the previous Core Update - aims to improve the search engine user experience. It is mainly about better matching results to the user's intentions or exploring the social and economic impact of chinese canadians in canada presenting them with a ready-made answer immediately in the search engine. The algorithm update should not be treated as a penalty for websites, and the decreases and increases result from changes intended to improve the quality of the service. For this reason, it is impossible to develop an effective strategy aimed at regaining lost positions. The only thing left is to continue to take care of the quality of the services and hope that the next update will be more gracious.

Image


Date of introduction: January 13, 2020
Purpose: To provide the search engine user with more relevant results.
Having trouble with your visibility in Google?

Rely on the specialists from KS!

Check out the offer!
May 2020 Core Update
SEOs have already gotten used to changes in Google algorithms. The implementation of the biggest ones is preceded by a message sent by Google SEO spokesman Danny Sullivan.

On May 4, 2020, he announced another major change to the search algorithm, which, in line with Google's latest actions, was named May 2020 Core Update. However, there is no information about what has actually been changed or improved. Google has been implementing major or minor improvements for the past several months that are supposed to improve the quality of use of their flagship service.

Algorithm changes are not always beneficial for sites that provide high-quality content to their readers. We have had many cases where one update of the main algorithm caused huge drops, and in the next one the traffic returned with a vengeance. Also after this update we should expect corrections that will at least partially compensate for the loss of traffic.

With every major reshuffle in search results, the question of how to react to changes in Google's search algorithms arises. There is no silver bullet, and even following Google's guidelines, taking care of technical aspects or content quality will not protect against a sudden drop resulting from an update.

Date introduced: May 4, 2020
Purpose: To provide the search engine user with more relevant results.
Summary
Google's algorithms are constantly evolving and adapting to changing conditions on the Internet. Just look at the updates regarding mobile display. If it weren't for the rapid development of smartphones and the ability to browse the Internet while traveling, such changes wouldn't be necessary.

It is impossible to list and describe all the updates in one article. There are simply too many of them, and they will grow over time. We can be sure of one thing. The algorithms that catalog pages will become more and more advanced and less susceptible to manipulation. Some of them will be (or have already been) adapted to learn on their own, which is why the ranking factors taken into account when arranging results in SERPs will constantly change.

I hope that this article has provided some insight into how the Google search engine works and what algorithms are responsible for the results you see after entering a specific phrase or question in the search window.