Surprising to see so much human input. Essentially google took a leaf out of Yahoo!'s playbook after first crushing them using just algorithms. Is this proof that beyond a certain degree of algorithmic extraction you need a human hand to advance or is it evidence for a return to the old days of a more curated index? (Or are there other possibilities?)
Google Releases the Full Version of Their Search Quality Rating Guidelines
11–20 of 22 posts
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#12Why this? Why not measure empirically based on how often people follow your first result?
I'm sure they measure that, also.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#13Why this? Why not measure empirically based on how often people follow your first result?
That approach is prone to creating a positive feedback loop through an order bias.
that's why they track what users do post-click. if the user quickly bounces back to the search results page , then the result probably wasn't a very good fit for that particular keyword.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#14Surprising to see so much human input. Essentially google took a leaf out of Yahoo!'s playbook after first crushing them using just algorithms. Is this proof that beyond a certain degree of algorithmic extraction you need a human hand to advance or is it evidence for a return to the old days of a more curated index? (Or are there other possibilities?)
Google has been upfront about greater use of learning AI (as opposed to hand-tuned algorithms like PageRank) to index and rank their search results. In order to return a good result, a learning AI system must first be trained on what a good result looks like. Since search returns results to humans, humans have to provide the guidance on what is a good or bad result.
Often an AI can be trained by a corpus, like training a spell correcter by feeding it the text of a dictionary. But the data that Google works is the entire web, which changes all the time, so they need constant feedback on what is good or bad.
Google also personalizes results now, so the AI needs to learn not just what is a good result, but what is a good result for that particular person, given everything else the AI knows about that person. That would require inputs from a wide variety of people.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#15Surprising to see so much human input. Essentially google took a leaf out of Yahoo!'s playbook after first crushing them using just algorithms. Is this proof that beyond a certain degree of algorithmic extraction you need a human hand to advance or is it evidence for a return to the old days of a more curated index? (Or are there other possibilities?)
I suspect this is also a fail safe which allows them to use learning algorithms that give less than well understood results that still perform well (think neural networks and deep learning).
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#16OT, the top rated comment on the above article published on googlewebmastercentral.blogspot.com is :
Google just made public their latest version of the Quality Raters Guidelines (PDF linked from the blog post)
Which is just summarizing what the blog post is about (probably a G+ share that got plus'd a lot). Nothing against the commentator, it's rather their ranking/voting system that needs to be fixed.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#17Document: http://static.googleusercontent.com/media/www.google.com/en/... Announcement: http://googlewebmastercentral.blogspot.com/2015/11/updating-...
Maybe dang can replace the link to the actual announcement and remove the blog spam.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#18The "your money or your life" web page stuff is interesting. They call these YMYL pages anything that can have large impact on the user: financial info pages, legal info pages, medical info pages, and so on. They say they have very high quality page rating standards for these pages. OK, but how? Later in the document they talk about doing reputation research on the site. They say, "[F]or Page Quality rating, you must…
From the announcement linked to below: ( https://news.ycombinator.com/item?id=10601232 ) Ratings from evaluators do not determine individual site rankings, but are used help us understand our experiments. The evaluators base their ratings on guidelines we give them; the guidelines reflect what Google thinks search users want
Ah so they probably have a specific set of sites that they check against. That makes sense- thanks.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#19Surprising to see so much human input. Essentially google took a leaf out of Yahoo!'s playbook after first crushing them using just algorithms. Is this proof that beyond a certain degree of algorithmic extraction you need a human hand to advance or is it evidence for a return to the old days of a more curated index? (Or are there other possibilities?)
I think it's proof that google has done an amazing sales job.
Their ML algos really come down to codified, scaled human intuition and decisions. However, to avoid having to answer for them (google has the ability to make or break many internet companies), they repeatedly go on and on about ML and pretend it's just math, as if there is a first principles answer to eg which candle company should rank highest passed down from Gauss.
The reason this is valuable to google is if they admit it is just collective human judgement, governments are much more likely to demand input.
Re: Google Releases the Full Version of Their Search Quality Rating Guidelines
#20Document: http://static.googleusercontent.com/media/www.google.com/en/... Announcement: http://googlewebmastercentral.blogspot.com/2015/11/updating-...
Maybe dang can replace the link to the actual announcement and remove the blog spam.
not downvoting you, but searchengineland is actually a very good/reputable site. It is very believable that the original submitter found this out through them. (IE: this isn't spam)