Connect with us

SEARCHENGINES

How BERT Helps With Google Search

Published

on

How BERT Helps With Google Search

Google has posted a new video on how BERT helps Google Search understand language. Google has been using BERT in search since 2018, we only knew about it in 2019. That being said, the short video basically says it is about Google understanding the little words better.

Here is the video:

Here is the transcript if you don’t want to listen:

If a pancake recipe told you to “mix the batter with the banana,” you probably wouldn’t think to use the banana as a mixing spoon. But what’s obvious to humans — things like context, tone, and intention — are actually very difficult for computers to pick up on. At its core, a Google Search is about understanding language. In order to return the right information, Google doesn’t just need to know the definition of the words… it needs to know what they all mean when strung together in a specific order. And that includes the smaller words like “for” and “to.” And when you think about how many different meanings a single word can have… you start to see how writing a computer program that takes all these nuances into account is pretty tough. See? Case in point. “Pretty” here doesn’t mean beautiful, it means “very.” More and more, people talk to Google the way they think and speak. And, more and more — Google is getting better at understanding what they mean. One of the biggest leaps forward in the history of Search came about with the introduction of “Bidirectional Encoder Representations from Transformers” or as we like to call it, BERT. BERT is a machine-learning, model architecture that helps Google process language and understand the context in which it appears. Search used to process a query by pulling out the words it thought were most important. For example, if you said, “can you get medicine for someone pharmacy” you would have gotten general results about pharmacies and prescriptions because it would have essentially ignored the word “for.” But with BERT, the LITTLE words are taken into account and it changes things. Search now understands you want to know if you can pick up medicine… prescribed to someone else. But how do you train a language model to pick up context? There’s a big difference between knowing words and understanding meaning. The model learns context by applying the same fill-in-the-blank principles it takes to complete a Mad Libs. So we take a phrase. We hide about 20% of the input words. And then we make the computer guess the words that are missing. Over time, the model begins to understand different words have different meanings depending on what’s around them. And the order in which they appear in that text, really matters. So when you search something complex like, “Fly fishing bait to use for trout in september montana” Search knows all the little words are important and because it now takes them all into account, Google can tell you the perfect bait for that time of year. BERT isn’t foolproof, but since implementing it in 2019, it’s improved a lot of searches. We should always be able to learn about whatever we’re curious about. And that’s why Search will always be working to understand exactly what you’re truly asking.

Advertisement

I am surprised Google did not release this video when we wrote about how Google uses AI in search.

Forum discussion at Twitter.

Advertisement




Source link

Keep an eye on what we are doing
Be the first to get latest updates and exclusive content straight to your email inbox.
We promise not to spam you. You can unsubscribe at any time.
Invalid email address

SEARCHENGINES

The Industry Mourns The Loss Of Mark Irvine

Published

on

Mark Irvine Rip

I am deeply sad to report that Mark Irvine passed away unexpectedly last night. Mark was deeply involved in the search marketing community, known as being a brilliant paid search specialist who always spent his time helping others over his long career in the space.

Mark has been in the search marketing industry for well over a decade, seven years at WordStream and the last four or so years at SearchLab. He was always a bright and welcoming smile at industry events, sharing advice with colleagues and friends.

Navah Hopkins worked with Mark for many years and she sent me the following message:

“Do less”

These were the words Mark would always say to me when we worked together. Yet he was incapable of doing less. He brought so much love and care to everything he did.

I was lucky enough to work with Mark directly for about 5 years, and counted him a friend since then. Beyond being a brilliant data scientist and a natural leader, Mark was one of the most human humans I’ve ever known. He understood people at a fundamental level and always made sure others around him could shine.

Advertisement

Mark would often say that I was his dedicated PR team and even as I’m writing this detailing his accomplishments and his amazing (and all together too short) life, I can hear him laughing that I’m doing it again. Well – tough…just enjoy the love we have for you and sip those heavenly pumpkin spice lattes.

Mark was always put data first behind every major project at WordStream, Navah told me. He was a worldly person who traveled abroad to speak at many events. He was really good at connecting with people abroad. Mark was named the #1 PPC influencer in 2019 and was consistently was on that list year-after-year. PPCHero has recognized him as one of the top PPC Influencers every year since 2015, as did Microsoft and in 2019.

Here is Mark’s bio at SearchLabs.

He leaves behind his partner, Bobby Main and his mother, Virginia Hall.

Here are some early reactions to the sad news of Mark passing:

Forum discussion at X.

Advertisement



Source link

Keep an eye on what we are doing
Be the first to get latest updates and exclusive content straight to your email inbox.
We promise not to spam you. You can unsubscribe at any time.
Invalid email address
Continue Reading

SEARCHENGINES

Daily Search Forum Recap: May 3, 2024

Published

on

Here is a recap of what happened in the search forums today, through the eyes of the Search Engine Roundtable and other search forums on the web.


Google got their cost down to generate AI answers in search by 80%. Google is interested in alternatives to hreflang. Google may recrawl URLs multiple times per day or per month. Google is testing custom search filter templates in search. Google AdSense removed its privacy policy as a placement for withdrawal of consent. And I posted the weekly SEO video recap. And deeply sad to report that passing of Mark Irvine.

Search Engine Roundtable Stories:

Other Great Search Threads:

Search Engine Land Stories:

Other Great Search Stories:

Analytics

Industry & Business

Links & Content Marketing

Local & Maps

Advertisement

Mobile & Voice

SEO

PPC

Other Search

Feedback:


Have feedback on this daily recap; let me know on Twitter @rustybrick or @seroundtable, on Threads, Mastodon and Bluesky and you can follow us on Facebook and on Google News and make sure to subscribe to the YouTube channel, Apple Podcasts, Spotify, Google Podcasts or just contact us the old fashion way.

Advertisement



Source link

Keep an eye on what we are doing
Be the first to get latest updates and exclusive content straight to your email inbox.
We promise not to spam you. You can unsubscribe at any time.
Invalid email address
Continue Reading

SEARCHENGINES

Google Interested In Hreflang Alternatives

Published

on

Google Pixel Globe

Gary Illyes from Google said he has heard from the SEO industry that hreflang can be “annoying” and complex and confusing. So he is open to ideas on how to replace it and make it work for both small and really large websites.

He wrote this on LinkedIn saying, “Things I’ve learned and heard in Sofia at the SERPConf event.” He said one of those is that “hreflang is annoying.”

Gary wrote, “I don’t disagree,” that I guess he understands why SEOs and creators find it annoying.

So he said that he open to new ideas. He wrote, “I’m still very open to coming up with something less annoying, but it needs to work for small sites and mammoths as well, while delivering at least the same amount of information.”

So if you have ideas, let Gary know, he wrote, “Ping if you have ideas.”

Advertisement

You can use hreflang to tell Google about the variations of your content. This helps Google understand the various pages and how they are localized variations of the same content. But the implementation can be confusing and detailed, why he said it can be annoying. Several years ago, John Mueller said hreflang can be the most complex aspect in SEO.

Forum discussion at LinkedIn.

Source link

Keep an eye on what we are doing
Be the first to get latest updates and exclusive content straight to your email inbox.
We promise not to spam you. You can unsubscribe at any time.
Invalid email address
Continue Reading

Trending

Follow by Email
RSS