Slashdot Mirror
Google's Mayer Says Personalization is Key To Future Search
rsmiller510 writes "In a wide-ranging interview with Tech Crunch's Michael Arrington on Wednesday at Le Web in Paris, Google's Marissa Mayer talked about all things Google, but what I found most interesting was when the conversation turned toward the future of search. Mayer said the key to the future of search lies in personalization. ... Mayer said in the future, Google (and presumably other search tools) will understand more about the user and be able to deliver more relevant information based on that knowledge. 'We think that when you look at the winning search engine in 2020 and what traits it's likely to have, we think the one thing that will be true is that it will understand more about you the user.'"
Video of the interview with Mayer is available at Tech Crunch. The personalization of search content focuses mainly on SearchWiki, which we discussed when it went live last month. The Register has a more cynical take on the discussion, seizing on comments by Mayer which indicated Google employees may evaluate SearchWiki's user ratings and use them to make "obvious changes" to search results for everyone.
Google will give you their web and email services, photo processing, mapping, and office applications and software that will make you a cup of tea. This is all paid for by personally-directed text ads in your tea leaves, based on analysing a DNA sample taken when you sip the tea and sending your genetic code back to Google for future targeting.
Google will maintain complete confidentiality within the marketing department of whatever they gather concerning your confidential business data, bank account details, medical information and personal preferences in pornography. "We're Google. We know where you live. In a completely not evil way. Sponsored link: Get Chrome Browsers on google.com. Or we'll make you use Windows Live. MuWAAAAhahaha. Sorry, that's my 'spreading Christmas cheer' laugh. Really."
http://rocknerd.co.uk
How convenient. The future of search requires giving up libraries of congress worth of personal info, to advertising companies with a sideline in search.
links with actual information. NOT links to sell me shit!
You, the Google Search/Gmail/Chrome user are the "product". The advertisers are the "customer". The servies on offer are the "bait". It's the television model.
http://rocknerd.co.uk
The future of Google is more data-mining of your private information. Fantastic. Please, bring on the bread and circus... I wonder when Google will delve into the "free" entertainment for your personal information business? GPorn? GPoker? GQuake? GWarcraft?
Light a fire for a man and he'll be warm for a day. Light a man on fire and he'll be warm for the rest of his life.
I've been working on a personalized news site, if anyone wants to see what the future looks like :) The site learns what you like (based on up/downvoting, like Reddit) and gives you only the news you want to see.
It would probably interest some Slashdot readers (it filters Slashdot, among other things, so you only see the good stuff) and it works like an RSS reader even if you don't trust the recommendations. It's at http://newsbrane.com/
While a Personalized search is likely to give you results that would probably most interest you, it is also likely to exclude results that don't fit your "profile" but are relevant to what you are looking for.
So the search engine is always returning results it thinks you would like to see, but not results that you probably should see.
When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
I don't want my search engine to know anything about me! For a number of reasons:
1. I don't want it applying pre-conceived notions of what I might be looking for. In some cases I'm sure it might be helpful for it to think it knows what I might want but I can see just as many cases where it would be bad as well.
2. I don't want them reporting, anymore than they already are, more stuff back to marketeers and such.
3. And even if the search engine does operate slightly better when it knows who I am that would mean it would require me to log in to it every place I go.
I can see that there could be advantages. But honestly the advantages look like they are a lot more on their end than mine.
Really, I know what I'm doing...Ohhhh, look at the shiny buttons!
You can use Google Search without having a Google account, you know. You don't even have to allow cookies or Javascript and you can block ads. You don't get all the neat personalized features, of course, but it works fine.
Warning: this article may contain humor, sarcasm, parody, and perhaps even irony. Read at your own risk.
I submitted this piece and while Mayer discussed the Search Wiki, this is not what I was referring to. If you fast forward to around the 16 or 17 minute mark of the video, Mayer begins a discussion of the future of search, 10-20 years out where she talks about personalization, modes and media as three pillars of future search. It is this part of the discussion that I focus on in my post. The Search Wiki actually has nothing to do with this.
That's the smartest thing that's been said here. Except the infrastructure of running superhuge data centers, plus the massive (both direct and indirect) networking, and the cost of having thousands of PhDs on staff is several orders of magnitude more expensive than for a TV network. Google sells a product, "consumer eyeballs" to customers (advertisers), anything they do to narrow the demographic "band" they can target for a particular ad makes their product more desirable.
No folly is more costly than the folly of intolerant idealism. - Winston Churchill
I think this can work OK if they gave you some way to edit your interests, or to temporarily turn them off. It would be really awful if it got stuck thinking you had some interest based on some odd search (or a misinterpretation of some term).
If I search for "gnome desktop" and "gimp" I would hate for it to start thinking I have a particular interest in men with unusual physical characteristics.
Amazon.com tracks what you search for and buy, and uses it to decide what ads to display. I like it, even though it makes me nervous to have them know too much about me. Amazon also gives you the option to say "I'm not really interested in that", so they can remove it from your list of interests.
Search personalization is of marginal value. In fact, it's kind of a pain, because searches become nonportable and nonrepeatable. If you tell someone else "search for ...", they won't get the same results you did. But advertising personalization... that's where the money is.
Google offers a great range of services and products, but almost all of them lose money.. No Google product other than search advertising makes money, and even that is declining. The Google Content Network (Google ads on non-Google sites) isn't that beneficial to the actual advertisers, and the more savvy advertisers have opted out of it. People click on those ads, but seldom buy. (By default, AdWords customers are opted in, and the opt-out checkbox is hard to find.) Google stock is down 57% from the peak, and revenue is projected to decline for the next three years. So Google is cutting back on new projects, killing off some of the money-losers, and trying to milk their one profitable product, ads on search results, for all they can.
Using search history, it would be straightforward to recognize specific big-ticket buying situations, like "looking for a car" or "looking for a house". This can be used for lead generation. Search for information about cars for a while, and not only do you start seeing car ads all the time, you get phone calls from sales reps.
People like you helping people like us help ourselves. - Processed World
Isn't the point of search engine is to know what other people think relevant regarding the subject that you're searching for?
The author of the Daniweb article fundamentally misunderstands semantic search, stating that "Search certainly has a key role in helping us sift through the mountains of information and semantic search, where the search engine has a sense of the meaning and context of our search, which is essentially what Mayer refers to when she cites personalization, could be the key to helping us access the data that's most relevant to us."
Actually semantic search is not really about personalisation at all. Personalisation could enhance semantic search, but they are two separate disciplines. I posted on what semantic search really means a while back, but I'll re-iterate here.
say a guitarist wants to search for a guitar amp speaker cabinet that would handle the 100w (that's RMS watts) output of his Marshall amp, and fit in the boot/trunk of his car - let's say it's a Ford whatever... anyhow, the point is that the semantic search app would need to discover the dimensions of the car's boot/trunk from Ford, then it could search for speaker cabs that would fit in the boot/trunk that could handle 100w (that's RMS remember) so, if the dimensions of the boot/trunk are expressed in inches by Ford and the dimensions of the cab are expressed in cm by Marshall/Mesa Boogie/Fender et al, then of course there needs to be a standard scheme to express units of size, and then the conversions are easy. Don't forget, that's 100w RMS...
Today you would need to run multiple searches, holding information (the size of the cabinet v.s. size of the boot/trunk) in your head to complete the task. With semantic search, the computer would do it all for you.
$ strings FTP.EXE | grep Copyright
@(#) Copyright (c) 1983 The Regents of the University of California.
The only issue I have with any kind of computer software that "learns" about your tendencies based on your past behavior is that human beings are unpredictable. We learn new things, we try new things and we aren't the same people today as we were 4 or 5 years ago. We grow and we change. So if I'm dealing with a search engine is that is filtering my results based on some previous information about my behavior what happens if I change?
Here's an extreme example. Let's say a drug user decides to clean up his act. He decides he wants to order a bunch of his favorite soft drink and have it delivered so he types in "coke". Well if he had previously searched for "coke" looking up information about the drug and how it affects your body and had previously been looking up information about other drugs and the search engine accounted for that then what comes up when he searches for "coke" but wants information about the soft drink? There better be an easy way for the system to set aside your previous behavior and start from scratch because people change and when you change the context in which you want information changes as well. Learning from past behavior has serious upside but there are some downsides as well that must be considered.
The future of search is not clogging every query result set with commercial links. You almost can't access legitimate information anymore, because every search returns advertisement after advertisement. It's almost to the point where you might as well go straight to wikipedia instead of bothering with Google.
Do what you can, with what you have, where you are.
For about a year or so I have been sending all my Googling through TOR (using Foxyproxy and a rule) as I do not like to be tracked, trended or advertised at, but it seems that Google are clamping down on people doing anonymous searches, under the old guise of blaming the end user and viruses etc. on their computer.
The problem is this. If you do a Google search through TOR, there is a very high chance you will get redirected to sorry.google.com and get a page back entitled "403 Forbidden" saying:
Google
Error
We're sorry... ... but your query looks similar to automated requests from a computer virus or spy ware application. To protect our users, we can't process your request right now.
We'll restore your access as quickly as possible, so try again soon. In the meantime, if you suspect that your computer or network has been infected, you might want to run a virus checker or spy ware remover to make sure that your systems are free of viruses and other spurious software.
If you're continually receiving this error, you may be able to resolve the problem by deleting your Google cookie and revisiting Google. For browser-specific instructions, please consult your browser's online support centre.
We apologise for the inconvenience and hope we'll see you again on Google.
When I started using Google through TOR it would work most of the time, with only the occasional one of these. Then the situation got better, as Google added a captcha to the page, so you could prove you were human and it would give you the results even though the freqency of the 403 went up. Then the capthca got rarer, and now I never see it, but get the 403 page near constantly. To work around the issue I have to tell my TOR daemon to create a new circuit, so I get a new exit node which might not be flooding Google as much. This obviously puts undue strain on the TOR network too.
The Google apologists will be queing up no doubt to tell me that I am getting this page because I am using TOR and all the searches appear to Google to come from a few IPs, so I should use Google directly. And whilst I am there I should forget about privacy and use a Google account all the time! All I can say is that Google's behaviour in reaction to searches from TOR is they appear to have made it harder for TOR users to use Google (do no evil, ha!), as 1) it used to work most of the time 2) TOR exit nodes are publicly known, so Google could easily whitelist those IPs 3) The captcha has gone away completely 4) the frequency of this error has shot up. I sometimes need to tell TOR to recreate the circuit over 5 times before I get an exit node that Google are serving.
I have also filled in feedback on the Google site on numerous occasions to try and get them to address the Googling through TOR issue, but I feel that they have just then clamped down on the untrackable people using TOR.
Google are very aggressive about trying to get info on their users, and it has now passed the point where they have decided that their business interests are important than people's anonymity. Google stopped being cool in my book a while ago, and these days they are utter bastards, who happen to operate the best search engine.
Car analogies break down.
Personalization is a red herring. A Google search is outward-facing, casting a line into the full depth and breadth of all of the world's information. I'm the one who gets to decide what is relevant to me.
No, I think Neal Stephenson nailed it in Anathem: the future of search is putting real value on reputation. After hundreds of years of global internet crap has accumulated in the world's indexes, supplemented by the output of intelligent marketing- and disinformation-bots, the only way to search with confidence will be to use a reputation market to filter the results. The results that providers are most willing to back with hard currency are the results you can put the most faith in.
I also wish Google would focus a little bit more on timeliness. It's really frustrating to search on an emerging Linux issue and get bombarded with results from 2003. It's a completely different OS now, that stuff is ancient history.
10 years on, and Google have a lot of data but are on the skids as a viable business. Imagine the US government offering to "help" by buying Google (and it's mountains of personal data) to "maintain stability". Imagine all Google's data being made available to every department of the US government for "national security" purposes; after all the war on terror will last a lifetime....right?
Many corporate events have happened which would have been unthinkable just a few years before the shit hit the fan. By the time it starts to crumble, people will start to wake up to the fact that Google knows a LOT more about them than they are comfy with.....specially if it can then be used by the government without any legal niceties like warrants, or even home / work / school visits.
I try to avoid Google as much as possible. I have done for a while, this story is just another reason not to change those habits.
Google is ridiculously more profitable than NBC Universal or CBS:
Google: http://finance.yahoo.com/q/is?s=GOOG&annual
NBC: http://www.ge.com/ar2007/mda_segop_summary.jsp
CBS: http://finance.yahoo.com/q/is?s=CBS&annual
NBC and CBS are far more than TV networks, so it doesn't directly address your point, but presumably, those companies integrated other aspects of the broadcasting business into their operations to save money, not to spend it.
That they provide more eyeballs at a lower price point is a huge aspect of the success that Google has had; I don't know, but I expect that being cheaper is part of those lower prices.
Nerd rage is the funniest rage.
Please tell me it didn't take this long for the million-dollar idea of GWARcraft to come about!
Is it wise for a single human to decide for his or her self what is and what is not "actual information"? It is a good idea to create tools that allow us to isolate ourself from mainstream thought? Maybe those "links that sell me shit" might mean "they have ads, but really the content is something that makes me uncomfortable". Who knows?
If the most widely used search engine filters the results so it returns only what you want to hear, what kind of society will we have? Will it create one where we humans debate ideas to find their core truths? Or will such technology allow us as a society to hide from uncomfortable truths because our tools make it easy?
Which do you prefer to live in?
I cannot be personalized.
My wish is less ambitious than that. The search engine and indexer need to have a basic understanding of the query and the page content. When I search for, say, 'image viewer for Mac', I want pages that are about -- astonishingly enough -- an image viewer for Mac, not a page that happens to have all those words scattered through it. When I search for pages about a person named 'John Lee', I want pages that at least mention a person with that name, not pages about 'John Lee Hooker', or about companies that happen to have 'John Lee' in their name. Enough semantics to know that 'photo' will do as a substitute for 'image', for example, would be helpful as well.
(You listening, Google? I'm an ontologist / NLP developer; let's talk.)