Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Search engines and SEO spam

twitter.com

Post reply on HN
Print view View original

181–190 of 555 posts

Re: Search engines and SEO spam

#181

Post by UncleMeat »

goatherders wrote: post #11
birken wrote: post #5

The funny thing is that if the people who worked on spam at Google were free to talk about it, I'm sure it would become evident that they know more about spam and anti-spam efforts than anybody else in existence. It's a ridiculously hard problem, especially when people are targeting you directly. But they aren't free to talk about it, because if they did it would just give more assistance to the spammers, and make th…

This x100000. There is no scenario - none - where thousands of engineers at Google working on search wake up in the morning and say "we sure have made it good enough wr2 SPAM. I think I'll have another Danish."

Especially since "made a change that improved search result relevance by X%" is an extremely compelling story for promotions. If indeed there is a launch-driven culture for promos at Google then there'd be extra incentive for new mechanisms to reduce low quality search results.

Re: Search engines and SEO spam

#182

Post by metabagel »

fragmede wrote: post #151
numbsafari wrote: post #117

Earlier quoted context omitted.

The problem isn't that Google doesn't employ these people or invest in their activities. It's that Google has destroyed their own search results in order to continue to expand their revenue opportunities. If Google: - Enabled downvoting on results, like YT videos. (Has its own spam problems, just like YT) - Allowed you to block certain domains from your search results, like YT videos. (If they added some kind of "coo…

How do you fight brigading, the organization of groups elsewhere to collectively vote on something? Eg white supremacist groups get together and vote down everything by people of color, and vote up their pages about how great they are?

This is a great point. I would think Google could rank users from low quality to high quality in terms of the quality of the websites which they recommend or downvote. Tricky business and could be difficult to control, but basically the same thing they currently do for websites, but extended to humans.

Re: Search engines and SEO spam

#183

Post by titzer »

PaulHoule wrote: post #170

For medical search the answer is pubmed. Not only is the collection of documents clean (of low-grade scammers, pharma companies have to pay big $ to play) but the NIH has done a large amount of search quality and ontology work -- the system knows "Tylenol" is synonymous with "Paracetamol", "Acetaminophen", etc.

> the system knows "Tylenol" is synonymous with "Paracetamol", "Acetaminophen", etc.

This is the exactly the kind of thing that Google cannot fathom manually doing. As if entering facts into a computer were morally wrong somehow. They'd much rather launch the equivalent of a shell script that harnesses face-melting amounts of computational power, processing literally trillions of webpages in bulk, signal and noise together, junk, spam, and misdirection alike, to learn bad associations and then serve them up with no human review and then put the full force of their reputation behind a results page that apparently people never check and certainly can't correct because of the inscrutability of a machine-learned model that has few to no levers to adjust.

Re: Search engines and SEO spam

#185

Post by streamofdigits »

Eventually search will become a decentralized activity (No, not a web3/crypto/coin type decentralization, I am talking about the useful type).

Is there any particular reason why internet search has to have a distorting gatekeeper to the global commons (that pretends playing Maxwell's demon). For chrissake, the stuff being indexed is public.

Re: Search engines and SEO spam

#186

Post by aronpye »

nikanj wrote: post #106
aronpye wrote: post #51

A lot of the spam results just seem to be copy pasted content. I wonder how difficult it is to compare the main body of text in search results, then say if it is over a 95% match with another site (I.e. it has been copy-pasted), demote it in the search results. If a site generates too many of these demotions then it gets blacklisted from the index.

How would you avoid throwing the original site out with the bathwater?

Maybe try and time stamp the page, presumably the earliest page is the original source. Could also combine it with a site reputation rating or something similar.

Re: Search engines and SEO spam

#187

Post by leoc »

https://twitter.com/mwseibel/status/1477707884632834049

> I’m pretty sure the engineers responsible for Google Search aren’t happy about the quality of results either. I’m wondering if this isn’t really a tech problem but the influence of some suit responsible for quarterly ad revenue increases.

Please no more of this. Two men, Page and Brin, together have basically unfettered control over Google.* If Google does something bad then, unless it's genuinely something small enough that those two could not be expected to hear about it, it's happening with—at the very least—their acquiescence. And low overall search quality is not something that some "suit" is successfully hiding from Good Czar Larry. They could fire the "suit", or command him or her to make other decisions. This is—again, at the very least—something that they have chosen not to do. The responsiblity lies with them.

* There is the risk of lawsuits from the minority shareholders, I assume. But IIUC this is not realistically that big a restraint on what shareholders with a majority of votes can do. However IANAL.

Re: Search engines and SEO spam

#188

Post by kevinmchugh »

awillen wrote: post #31
judge2020 wrote: post #15

Earlier quoted context omitted.

Also, i'd be very surprised if they didn't have tens of thousands of workers aiding in spam review already. The hard part in all of this isn't finding and stopping spam - it's defining what spam is. Are all the pie recipes where there's a 2000 word essay about their grandma at the top 'spam'? They still have the recipe, and Google Home devices pick up the recipe instructions just fine so people end up not reading it,…

AFAIK the 2000 word essays in recipes are Google's fault - it prioritizes pages with a lot of content, so you have to add that junk to the top in order to rank highly. While I'm sure there's more going on behind the scenes than I'm aware of, it does seem like the rules could be altered on a category-specific basis where a lot of text isn't necessarily a positive.

Recipe intro text is useful for contextualizing the recipe and copyright purposes. In RSS days, it was a way to get readers to click through, so the author got the ad views. Also people who write recipes like to write about food.

Re: Search engines and SEO spam

#190

Post by ocdtrekkie »

birken wrote: post #5

The funny thing is that if the people who worked on spam at Google were free to talk about it, I'm sure it would become evident that they know more about spam and anti-spam efforts than anybody else in existence. It's a ridiculously hard problem, especially when people are targeting you directly. But they aren't free to talk about it, because if they did it would just give more assistance to the spammers, and make th…

I think the problem is just that the solution isn't in Google's wheelhouse: There is no algorithmic ranking system that can't be gamed. Human moderation and curation is the only way to provide true quality, and Google is allergic to solutions that don't automate and scale.

I think a really good search engine would still algorithmically search it's index, but the content library should be human-curated with a goal of ingesting content via author, not via platform. Once a given author was human-approved as a quality source of information, content they produce could be automatically ingested going forwards, and conditionally re-reviewed by a human if there were reports the quality had decreased.

Post reply on HN