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

Facebook machine learning technology improves; Redditors alaramed.

reddit.com

Post reply on HN
Print view View original

151–160 of 168 posts

Re: Facebook machine learning technology improves; Redditors alaramed.

#151

Post by jemfinch »

falling wrote: post #150
jemfinch wrote: post #142

Earlier quoted context omitted.

> However, I think that if that 1 service makes it possible for you to backup all your information and allows you to remove all the information from their service if you want to, then it shouldn't be a problem... Something like Google's Data Liberation Front? http://www.dataliberation.org/ I put all my eggs in one basket every time I take the family to the grocery store. It's unclear to me why storing all my informat…

Facebook lets you download your data as well: https://www.facebook.com/help/?topic=download Just wanted to balance the “Facebook evil / Google not evil” diatribe, as when discussing these topics only Google is ever mentioned.

How long has this existed? I don't recall seeing such an option when I deactivated my account almost a year ago, and it might be worth trying to reactivate my account just to get this data.

Re: Facebook machine learning technology improves; Redditors alaramed.

#152

Post by falling »

jemfinch wrote: post #151
falling wrote: post #150

Earlier quoted context omitted.

Facebook lets you download your data as well: https://www.facebook.com/help/?topic=download Just wanted to balance the “Facebook evil / Google not evil” diatribe, as when discussing these topics only Google is ever mentioned.

How long has this existed? I don't recall seeing such an option when I deactivated my account almost a year ago, and it might be worth trying to reactivate my account just to get this data.

Most of the news articles talking about this are from early October 2010, about the same time the Google Data Liberation Front was created.

Re: Facebook machine learning technology improves; Redditors alaramed.

#153

Post by patrickaljord »

falling wrote: post #150
jemfinch wrote: post #142

Earlier quoted context omitted.

> However, I think that if that 1 service makes it possible for you to backup all your information and allows you to remove all the information from their service if you want to, then it shouldn't be a problem... Something like Google's Data Liberation Front? http://www.dataliberation.org/ I put all my eggs in one basket every time I take the family to the grocery store. It's unclear to me why storing all my informat…

Facebook lets you download your data as well: https://www.facebook.com/help/?topic=download Just wanted to balance the “Facebook evil / Google not evil” diatribe, as when discussing these topics only Google is ever mentioned.

Google allows you to download your data, which includes the emails of your contacts, your icals etc. In Facebook you get your list of contacts consisting in a list of full names in pure text, no emails. The only way to get in contact with them is to connect with them on facebook, making the data completely useless. With Google, you can export everything and use it somewhere else and still be in touch with your contacts, calendars etc _without_ ever using google again. Big difference.

Re: Facebook machine learning technology improves; Redditors alaramed.

#154

Post by nandemo »

emw wrote: post #114
sondh wrote: post #50

I don't think this is machine learning. I have Facebook suggested a few albums today (2 hours ago). Some of them are from 2+ years ago, others are recent ones. The suggestion is quite accurate (89%, 8 out of 9 albums). With the incorrect album, Facebook suggested another city with the same name but from another country! So Facebook probably just combine as much data as possible and when the matching rate is larger th…

I second the idea that this doesn't sound like machine learning. I wouldn't be surprised if there were no machine learning or data mining algorithms involved in Facebook mapping EXIF-embedded geographic coordinates in digital photographs and suggesting locations for an album based on points-of-interest near those coordinates. Maybe some neat work with GIS enabled this, but I don't see anything that strongly indicates…

The reddit poster claimed they didn't have GPS on their camera. What happened is problaby (3) above:

http://www.reddit.com/r/WTF/comments/nkktm/facebook_is_reall...

Re: Facebook machine learning technology improves; Redditors alaramed.

#155

Post by henrybridge »

iamandrus wrote: post #146
brlewis wrote: post #138

This was just answered on Quora: http://www.quora.com/How-does-Facebook-predict-my-photo-albu... Henry F. Bridge, Product Manager, Facebook Photo albums on Facebook have long had a text input field for location, but until recently, there was no way to put in structured data in this field (like a Facebook page). We added that ability earlier this year, and the "add a location" feature just performs a search on whateve…

I uploaded pictures from a class trip on Facebook back in 2009 and had no captions or names that suggested that the pictures were taken in California. My camera at the time was not capable of GPS. Facebook still managed to get the location right. Scary stuff.

Like I said on Quora, we use the location field that you put into the album when you uploaded it. The album location isn't displayed in the caption for individual photos and isn't displayed that prominently on the album page, so you're probably just not seeing that you set it when it was uploaded in 2009.

Re: Facebook machine learning technology improves; Redditors alaramed.

#157

Post by apu »

drats wrote: post #24

http://en.wikipedia.org/wiki/Scale-invariant_feature_transfo... Facebook almost certainly has more photo information than TinEye or Flickr, and of indoor environments probably more than Google (which has reverse image search too). Across any given bar or hospital Facebook would have maybe 5-10 other people with albums tagged with the name/gps/check-in. They'd only need one other album though. SIFT more or less turns…

SIFT is not enough for this task. There are many computer vision researchers working on this application (automatically inferring location of an image), and accuracies are pretty low right now, except for popular landmarks.

The problem is that single SIFT features are not very distinctive, and so you need many of them in common between two images to get reliable matches. So you essentially need images taken from very similar locations. You can ameliorate this requirement slightly by using fancy tricks in how you aggregate SIFT features together, but the fundamental constraint remains.

This precondition is satisfied at popular landmarks, since there is a good chance that someone's taken a photo from the same location you're standing at, but in general, this condition is not that easy to meet, and hence the poor accuracy of current approaches.

Finally, SIFT is great only for roughly-2d, distinctively textured areas. However, a lot (maybe most?) of the world does NOT fall into this category -- many things are either not textured, not distinctive, or are 3d. Something other than local descriptors (of which SIFT is the best known example) will be needed to understand these kind of scenes.

If you're curious about this line of research, a good place to start is the IM2GPS paper, which was the first major work (that I'm aware of) to look at this problem: http://graphics.cs.cmu.edu/projects/im2gps/

Re: Facebook machine learning technology improves; Redditors alaramed.

#158

Post by iamandrus »

henrybridge wrote: post #155
iamandrus wrote: post #146

Earlier quoted context omitted.

I uploaded pictures from a class trip on Facebook back in 2009 and had no captions or names that suggested that the pictures were taken in California. My camera at the time was not capable of GPS. Facebook still managed to get the location right. Scary stuff.

Like I said on Quora, we use the location field that you put into the album when you uploaded it. The album location isn't displayed in the caption for individual photos and isn't displayed that prominently on the album page, so you're probably just not seeing that you set it when it was uploaded in 2009.

I never entered anything into the location field (I would have remembered doing it), yet it got the exact city and state where I took the photos.

Re: Facebook machine learning technology improves; Redditors alaramed.

#159

Post by sajidnizami »

There was an old vision that machines would do the stuff for you somewhere back in the day.

Sad that we've got to a point where we can make this possible but won't because we can't trust pie in the sky.

Torn between the two options and can't decide!

Re: Facebook machine learning technology improves; Redditors alaramed.

#160

Post by rhizome »

MrRage wrote: post #145
nitrogen wrote: post #124

Earlier quoted context omitted.

While people like to think that it's more or less anonymized clusters and demographics that are being passed to advertisers to target ads better - not a terribly scary thought, just more relevant ads - I think we need to be far more cautious. How do you go about convincing the "I have nothing to hide" majority of the population that giving away such vast quantities of information is not in their best interests? I'm t…

I'm not sure how compelling my argument is... but if I have nothing to hide then why does anybody need to look? For example does the TSA seeing everybody naked actually improve security?

Everybody has something to hide. If you don't understand what I mean by this, post your credit card number in a reply.
Post reply on HN