You can't just have a one line write up. Who the hell is Lori Drew? Is it really that hard to ad the line: "Lori Drew is the woman who was convicted of using MySpace to tease a 13 year old girl until the girl committed suicide."
I believe that Netflix is still using Cinematch. You could look into movielens. It's from the GroupLens group at U Minn.
[E]ven though both probably use a pretty simple algorithm compared to Nextflix.
You do know that Netflix said on the outset "You're competing with 15 years of really smart people banging away at the problem." and it was beat in less than week.
That's not to meant as a knock against Netflix's engineers, but more about that they didn't really build a state of the art recommender system. Simple SVD (aka latent semantic indexing outperfomed them as well.) They did something a bit more than straight up kNN clustering, but that was pretty much it.
1.) Rare could also be defined as unpopular. Trying to recommend unpopular movies is problematic. Is the computer program going to be able to discern under-rated (Glengarry Glen Ross) or just crap (Ishtar)
You know what. I actually like Ishtar. I really do. The blind camel, and the line "We're not singers! We're songwriters!" gets me every time.
So really, the even harder problem is to know when to buck your your friends and go with the the outlier. It's hard, because kNN methods work pretty well, and they're all about going with the consensus of whatever cluster you're in.
I published a paper using Netflix data. (Yeah, that group.)
It's certainly cool that they beat the 10% improvement, and it's a hell of a deal for Netflix, since it would have cost them more than a prize money paid out to hire the researchers, the interesting thing is whether or not this really advances the the field of recommendation systems.
The initial work definitely did, but I wonder how much of the quest for the 10% threshold moved the science, as opposed to just tweaking an application. Recommender systems still don't bring up rare items, and they still have problems with diversity. None of the Netflix Prize work address any of these problems.
How *does* one get a Masters Degree in Information Science without being able to answer basic questions like this?
Easy. Information Science is basically "Clicking the B in the toolbar makes the text bold." You're problem is that you're not interviewing computer science majors.
Wait. So some guy in the US with a twitter account is attacked by what is definitely implied to be Basij militia, and it makes news on some buried diary on a political activist blog? What? The guy didn't have a guts to carve an A into his own cheek? (At least an 'A' is symmetric.)
Wait. Claiming audio sucks on Linux is FUD because there's not one, not two, but three mutually incompatible and redundant APIs? How the hell is this not a clusterfuck?
Oh I'm sure there's some reason why someone prefers one to the other, but seriously. You're sending bits to a soundcard. That's it. Just make one API and be done with it. Got a beef with the API? Enhance it, don't just throw it away?
My god, audio was one of the reasons why I ditched Linux for a mac four years ago after running it as my primary OS for ten years prior. Frankly I got tired of having sound work in some applications, but not others. I got tired of guessing which mixer would adjust the sound, which mixer wouldn't. I got tired of seeing "No ALSA cards detected" in my startup, but someone how having `alsamixer` be the one mixer that worked most consistently.
This is a mess made by the developer community and developer community has so far failed to show that it is capable of solving it. If only there were a Benevolent Dictator or something...
Perhaps I wasn't sarcastic enough... Yes, I am aware they are one and the same now (altavista provides search for Yahoo).
Eh hem. Altavista ceased to exist in any way beyond a search box used for random beta testing years ago. Altavista got bought by Overture, which made ads. Specifically they mad a keyword auction. Overture got bought by Yahoo back in like 2005 or 2006. You may recall that Yahoo had a search engine back then. Who made it? Here's a hint: They provided search to MSN, Yahoo, and HotBot (Remember them? No. Of course not whippersnapper.) back in The Day(tm), until they were acquired by Yahoo and turned into Yahoo Search Technology. Who was this? Inktomi.
Frankly,looking over the constitutional powers allotted to the federal government they have no f**king business buying businesses, funding research, baling businesses out, or a large host of other "responsibilities" they have taken on illegally.
Q: Where does it say that the federal government can't do this? A: It doesn't.
Q: When has the federal government ever behaved they way you think it should? A: Never.
They are supposed to collect tariffs on imports rather than tax the citizenry.
AT&T is heavily investing into their infrastructure as the market continues to demand more.
Bull. Shit. This is the same crap that AT&T and all the other telcos spout, but at the same time, the US still has one of the worst cell systems in the industrialized world. We still have one of the worst deployment of high speed Internet connectivity. And don't even try that "but the population is denser there" argument. You only have to wire up where the people live. No one is asking you to wire up Denali.
Of course, the name does enormous things for your placement in google. Just do a google search for "buy flowers": at least half the results have the search the search terms right in the domain name. This is not a coincidence. If the name describes what you do and is also your branded name, your success in google is almost guaranteed.
Keyword in domain is just one of literally hundreds of features that go into ranking function. Let's take your "buy flowers" search. The top hits are:
1800flowers.com
buyflowersonline.com
ftd.com
buyflowers.net
honestflorist.com
beyondblossoms.com
flowershopdeals.com
buyflowers.org
proflowers.com
onlineflowers.com
Looking at the top-5 (which is the ones anyone really ever see), one has one keyword (the top hit), two have both, and two have neither. If we take the entire top-10, we still only have three with both keywords, six with one, and three with none. So what can we conclude about this? First, it seems pretty obvious that the "buy" keyword in the domain isn't very useful, otherwise we'd have a lot more "buy foo" hits, however the "flowers" keyword seems pretty good. So why is that? Well if you're selling flowers, it certainly seems natural to put "flowers" in the name of your company, and then use your company name for your domain.
I'm on/. with a UID an order of magnitude smaller than yours. I know the difference. Still, I stand by my comment. "As a nerd, and also a comp sci major" you should know that algorithms and implementations are not the same, and in fact implementation details are irrelevant when compared with to algorithm details. By divorcing the description of the algorithm from the implementation it's not only easier to understand the algorithm, but also to focus on how to improve the algorithm. By giving the implementation in the form of a playable game not only are you losing your income, but all you're doing is encouraging people to do is tweak the implementation rather than algorithm. (e.g. "If you were to use an array instead of a vector, your bubble sort would run much faster.")
The question is whether the Chinese raised in the communist propaganda can handle the criticism of their own government without taking it at the personal level and getting all emotional and defensive.
When I had a discussion with an expat about how the Chinese propaganda I saw about it said that the PLA installed "democratic reforms" in Tibet after the 1950 invasion, and how I said that was the most glaringly obvious lie (as opposed to having an element of truth) he proceeded to attack the US and brought up The Bonus Army(!). Wow! Nothing violent happened at Tiananmen Square, and anyway if it did. the US is just as guilty because of the MacArthur shot the Bonus Army. My Chinese friend was dumbfounded when I responded, with "Yeah. I know about that. The veterans didn't get paid, and then during their protest they got shot. It shouldn't have happened. It was wrong. What's your point?" Honestly, it was like his mom never told him, "Just because everyone else does it doesn't make it right."
I wonder how feasible it would be for the Internet crowd to "make" June 4 the unofficial day of the free speech, by means of posting some small banner or a short comment on thousands of websites on that day, to the extent that it would get media coverage, and then repeating it every year on the anniversary of the Tiananmen massacre.
Here's the thing though. Putting up a banner of Tank Man on your blog doesn't matter. No one in China reads it. Putting Tank Man as the background to Google doesn't matter. No one in China visits it. They use Baidu. My point is, that you have to put your message where your audience is, and the audience is predominately visits the.cn domain, if for no other reason than the convenience of the language and culture. Same reason why you don't visit the Chinese US-expat site mitbbs.com. It's the wrong language, and they don't talk about anything you really care about.
So let's you managed to pull this off. You've used enough sites that are outside of the reach of Jingjing and Chacha, that word has spread though the mainland Chinese community. What's the reaction? I suspect that it would be just like the Olympic torch protests. anti-cnn.com whips up some nationalist fervor about how the West is jealous China's progress and are just trying to tear it down. A billion MSN users all add "(L)CHINA" back to their statuses.
those who came from China to study and may be oblivious of the fact that the rest of the world considers Chinese government's policies and actions morally questionable.
Have you talked to any recent Chinese expats? I have. It can be fucking twisted. I was talking with one about China and how the US perceives it. (I visited Beijing for two weeks before the Olympics.) I told him that it was obvious that today's China is not Mao's China. It's not a palatable dysfunctional police state like North Korea. Obviously China has come far economically in a very short time. However, it is still clear that the old guard still exists. (As the Olympic preparations sign said, "When meeting foreigners remember these simple rules: #1 Maintain the social harmony...") I told him about the English language "documentary" about 1950 Tibet Invasion...err..."Invitation" and how it repeatedly talked about how the PLA brought (I swear to the god your choice, this is a direct quote) "democratic reforms" to Tibet. No. Not a chance. Not Chairman Mao. Not in a million years. Yes, Tibet was a lot less Shangri-la than a lot of Westerns want to believe, but there's no way that the Chairman Mao brought democracy to anyone. The expat's response? "Well, they made it more democratic." "No. It doesn't even meet the dictionary definition of a democracy. Not even a crooked one." "But they made it MORE democratic. You sound like they just changed kings." "They did." "No. I don't believe that." Then he said why democracy hadn't come to China. "It is. Just very slowly. We don't want it end up like Russia." "Well, that was an economic collapse, and you've avoided that. Plus that doesn't explain how Germany integrated, or Poland, or the Czech Republic, or Romania, or really any of Eastern Europe with the exception of the former Yugoslavia."
That said, while I was in Beijing, I had a similar long conversation with a friend that lives there, and it was different.
That's why you have a second xterm open with `make |& less` constantly queued up on the command line. Seriously. There's nothing an IDE does that you can't do faster and easier. Learn to write a makefile and an editor and your skills are portable. Learn an IDE and you're screwed as soon as you change jobs and they say, "Oh. We're a foo shop."
You should also be using IDEs and editors that support multiple languages.
The vast majority of IDEs are stuck for exactly language, or perhaps two. You want something that handles everything. Use an editor. If you use a grownup editor like vi or emacs, you've got syntactic markup for pretty much everything, along with language specific folding. IDEs simply don't do that. Not that they can't, but they don't. It's because the IDE designers say, "Wouldn't it be cool if we built an IDE for foo?", and the IDE is biased towards that language and workflow.
Gone are the days when a computer professional could afford to know just one or two languages. Who the hell wants to learn the quirks of different editors for each one?
So we should learn the quirks of behemoth IDEs instead?
Seriously. Why bother? IDEs are so hard to simply so hard to setup. Whenever I try to use one, I always end up fighting with trying to add nonstandard libraries, or hooking it into the code repository. It's horrible. Seriously, a three line makefile is all you need. It is so much easier just being able to say, "You! Compile this, with this option." It's 50 thousand clicks and it still doesn't work.
It seem like the obvious approach would be a fairly large RC plane and mount a second the camera (perhaps on a servo) and a tv transmitter on it. You downlink the video to a laptop that then uses some sort of usb connection to a gutted rc controller, either with servos moving the sticks directly, or better yet, bypassing the potentiometers and variably outputting voltage directly to the control board.
It seems like the hardest thing is avoiding (auto)pilot error. I don't have any experience with RC planes, but from what I've heard you have to go into with the attitude that you're going to spend a thousand dollars for 10 seconds of entertainment. You just have to assume that the plane is going to be destroyed on its first flight. Anything after that is bonus.
1. Why are you reinventing the wheel? 2. What makes you think that anyone wants to look at this? 3. Why do you want to waste your time babysitting a bunch of technology on vacation?
Usually when the government lowers taxes they see an increase in tax revenue because of increased spending since taxes are lower.
The "usually" here is a wonderful weasel word. You can't keep cutting taxes and seeing an increase in revenue. I mean think about it. If it was always true, why wouldn't you just cut your tax rate to zero and to maximize tax revenue? It doesn't make any sense. There's a point where you are simply cutting into revenue. Similarly, raising taxes will increase revenue, but there comes a point where they drag down the entire economy, thus cutting into revenues. There's a term for this, it's called "diminishing returns."
The fact is, that in the US the tax rate is one of the lowest of western world. In fact, according to the conservative American Enterprise Institute, tax revenue actually dropped as a result of the Bush tax cuts. The nonpartisan Congressional Budget Office came to the same conclusion in 2007.
Economies, especially modern economies, are complex interconnected systems. That means there are no simple solutions, and there are definitely no cure-alls. Those that constantly advocate tax cuts without regard to the situation or an understanding of their effects in the current economy, are simply advocating a means instead of ends. It's zealotry, and demonstrates no understanding of the problem.
Two guesses how that will turn out.... Not like it hasn't been tried before.
I'll go with, "better than the tax cuts that decreased revenue."
You can't just have a one line write up. Who the hell is Lori Drew? Is it really that hard to ad the line: "Lori Drew is the woman who was convicted of using MySpace to tease a 13 year old girl until the girl committed suicide."
Apparently it is.
I believe that Netflix is still using Cinematch. You could look into movielens. It's from the GroupLens group at U Minn.
You do know that Netflix said on the outset "You're competing with 15 years of really smart people banging away at the problem." and it was beat in less than week.
That's not to meant as a knock against Netflix's engineers, but more about that they didn't really build a state of the art recommender system. Simple SVD (aka latent semantic indexing outperfomed them as well.) They did something a bit more than straight up kNN clustering, but that was pretty much it.
You know what. I actually like Ishtar. I really do. The blind camel, and the line "We're not singers! We're songwriters!" gets me every time.
So really, the even harder problem is to know when to buck your your friends and go with the the outlier. It's hard, because kNN methods work pretty well, and they're all about going with the consensus of whatever cluster you're in.
I published a paper using Netflix data. (Yeah, that group.)
It's certainly cool that they beat the 10% improvement, and it's a hell of a deal for Netflix, since it would have cost them more than a prize money paid out to hire the researchers, the interesting thing is whether or not this really advances the the field of recommendation systems.
The initial work definitely did, but I wonder how much of the quest for the 10% threshold moved the science, as opposed to just tweaking an application. Recommender systems still don't bring up rare items, and they still have problems with diversity. None of the Netflix Prize work address any of these problems.
Still, I look forward to their paper.
Easy. Information Science is basically "Clicking the B in the toolbar makes the text bold." You're problem is that you're not interviewing computer science majors.
Yes. Is it his fault? No, it's United States' lack of universal care like every other country in the world has.
Wait. So some guy in the US with a twitter account is attacked by what is definitely implied to be Basij militia, and it makes news on some buried diary on a political activist blog? What? The guy didn't have a guts to carve an A into his own cheek? (At least an 'A' is symmetric.)
Wait. Claiming audio sucks on Linux is FUD because there's not one, not two, but three mutually incompatible and redundant APIs? How the hell is this not a clusterfuck?
Oh I'm sure there's some reason why someone prefers one to the other, but seriously. You're sending bits to a soundcard. That's it. Just make one API and be done with it. Got a beef with the API? Enhance it, don't just throw it away?
My god, audio was one of the reasons why I ditched Linux for a mac four years ago after running it as my primary OS for ten years prior. Frankly I got tired of having sound work in some applications, but not others. I got tired of guessing which mixer would adjust the sound, which mixer wouldn't. I got tired of seeing "No ALSA cards detected" in my startup, but someone how having `alsamixer` be the one mixer that worked most consistently.
This is a mess made by the developer community and developer community has so far failed to show that it is capable of solving it. If only there were a Benevolent Dictator or something...
Eh hem. Altavista ceased to exist in any way beyond a search box used for random beta testing years ago. Altavista got bought by Overture, which made ads. Specifically they mad a keyword auction. Overture got bought by Yahoo back in like 2005 or 2006. You may recall that Yahoo had a search engine back then. Who made it? Here's a hint: They provided search to MSN, Yahoo, and HotBot (Remember them? No. Of course not whippersnapper.) back in The Day(tm), until they were acquired by Yahoo and turned into Yahoo Search Technology. Who was this? Inktomi.
Q: Where does it say that the federal government can't do this?
A: It doesn't.
Q: When has the federal government ever behaved they way you think it should?
A: Never.
Good thing the 16th Amendment doesn't exist.
Wouldn't a "wyminist" be a misogynist?
YHL.
Finally, something that can not only see Russia, but also defend us when Putin raises his head over the airspace in Alaska.
Bull. Shit. This is the same crap that AT&T and all the other telcos spout, but at the same time, the US still has one of the worst cell systems in the industrialized world. We still have one of the worst deployment of high speed Internet connectivity. And don't even try that "but the population is denser there" argument. You only have to wire up where the people live. No one is asking you to wire up Denali.
Arbitrage is not "adding value".
Keyword in domain is just one of literally hundreds of features that go into ranking function. Let's take your "buy flowers" search. The top hits are:
Looking at the top-5 (which is the ones anyone really ever see), one has one keyword (the top hit), two have both, and two have neither. If we take the entire top-10, we still only have three with both keywords, six with one, and three with none. So what can we conclude about this? First, it seems pretty obvious that the "buy" keyword in the domain isn't very useful, otherwise we'd have a lot more "buy foo" hits, however the "flowers" keyword seems pretty good. So why is that? Well if you're selling flowers, it certainly seems natural to put "flowers" in the name of your company, and then use your company name for your domain.
I'm on /. with a UID an order of magnitude smaller than yours. I know the difference. Still, I stand by my comment. "As a nerd, and also a comp sci major" you should know that algorithms and implementations are not the same, and in fact implementation details are irrelevant when compared with to algorithm details. By divorcing the description of the algorithm from the implementation it's not only easier to understand the algorithm, but also to focus on how to improve the algorithm. By giving the implementation in the form of a playable game not only are you losing your income, but all you're doing is encouraging people to do is tweak the implementation rather than algorithm. (e.g. "If you were to use an array instead of a vector, your bubble sort would run much faster.")
Yeah, because the techniques are completely unreproducible. Ideas translate implementations. You just want a free game.
The question is whether the Chinese raised in the communist propaganda can handle the criticism of their own government without taking it at the personal level and getting all emotional and defensive.
When I had a discussion with an expat about how the Chinese propaganda I saw about it said that the PLA installed "democratic reforms" in Tibet after the 1950 invasion, and how I said that was the most glaringly obvious lie (as opposed to having an element of truth) he proceeded to attack the US and brought up The Bonus Army(!). Wow! Nothing violent happened at Tiananmen Square, and anyway if it did. the US is just as guilty because of the MacArthur shot the Bonus Army. My Chinese friend was dumbfounded when I responded, with "Yeah. I know about that. The veterans didn't get paid, and then during their protest they got shot. It shouldn't have happened. It was wrong. What's your point?" Honestly, it was like his mom never told him, "Just because everyone else does it doesn't make it right."
Here's the thing though. Putting up a banner of Tank Man on your blog doesn't matter. No one in China reads it. Putting Tank Man as the background to Google doesn't matter. No one in China visits it. They use Baidu. My point is, that you have to put your message where your audience is, and the audience is predominately visits the .cn domain, if for no other reason than the convenience of the language and culture. Same reason why you don't visit the Chinese US-expat site mitbbs.com. It's the wrong language, and they don't talk about anything you really care about.
So let's you managed to pull this off. You've used enough sites that are outside of the reach of Jingjing and Chacha, that word has spread though the mainland Chinese community. What's the reaction? I suspect that it would be just like the Olympic torch protests. anti-cnn.com whips up some nationalist fervor about how the West is jealous China's progress and are just trying to tear it down. A billion MSN users all add "(L)CHINA" back to their statuses.
those who came from China to study and may be oblivious of the fact that the rest of the world considers Chinese government's policies and actions morally questionable.
Have you talked to any recent Chinese expats? I have. It can be fucking twisted. I was talking with one about China and how the US perceives it. (I visited Beijing for two weeks before the Olympics.) I told him that it was obvious that today's China is not Mao's China. It's not a palatable dysfunctional police state like North Korea. Obviously China has come far economically in a very short time. However, it is still clear that the old guard still exists. (As the Olympic preparations sign said, "When meeting foreigners remember these simple rules: #1 Maintain the social harmony...") I told him about the English language "documentary" about 1950 Tibet Invasion...err..."Invitation" and how it repeatedly talked about how the PLA brought (I swear to the god your choice, this is a direct quote) "democratic reforms" to Tibet. No. Not a chance. Not Chairman Mao. Not in a million years. Yes, Tibet was a lot less Shangri-la than a lot of Westerns want to believe, but there's no way that the Chairman Mao brought democracy to anyone. The expat's response? "Well, they made it more democratic." "No. It doesn't even meet the dictionary definition of a democracy. Not even a crooked one." "But they made it MORE democratic. You sound like they just changed kings." "They did." "No. I don't believe that." Then he said why democracy hadn't come to China. "It is. Just very slowly. We don't want it end up like Russia." "Well, that was an economic collapse, and you've avoided that. Plus that doesn't explain how Germany integrated, or Poland, or the Czech Republic, or Romania, or really any of Eastern Europe with the exception of the former Yugoslavia."
That said, while I was in Beijing, I had a similar long conversation with a friend that lives there, and it was different.
Dude. What are you doing? Why did you have to show me up like that in front of everyone?
That's why you have a second xterm open with `make |& less` constantly queued up on the command line. Seriously. There's nothing an IDE does that you can't do faster and easier. Learn to write a makefile and an editor and your skills are portable. Learn an IDE and you're screwed as soon as you change jobs and they say, "Oh. We're a foo shop."
The vast majority of IDEs are stuck for exactly language, or perhaps two. You want something that handles everything. Use an editor. If you use a grownup editor like vi or emacs, you've got syntactic markup for pretty much everything, along with language specific folding. IDEs simply don't do that. Not that they can't, but they don't. It's because the IDE designers say, "Wouldn't it be cool if we built an IDE for foo?", and the IDE is biased towards that language and workflow.
So we should learn the quirks of behemoth IDEs instead?
Seriously. Why bother? IDEs are so hard to simply so hard to setup. Whenever I try to use one, I always end up fighting with trying to add nonstandard libraries, or hooking it into the code repository. It's horrible. Seriously, a three line makefile is all you need. It is so much easier just being able to say, "You! Compile this, with this option." It's 50 thousand clicks and it still doesn't work.
God, IDEs suck.
It seem like the obvious approach would be a fairly large RC plane and mount a second the camera (perhaps on a servo) and a tv transmitter on it. You downlink the video to a laptop that then uses some sort of usb connection to a gutted rc controller, either with servos moving the sticks directly, or better yet, bypassing the potentiometers and variably outputting voltage directly to the control board.
It seems like the hardest thing is avoiding (auto)pilot error. I don't have any experience with RC planes, but from what I've heard you have to go into with the attitude that you're going to spend a thousand dollars for 10 seconds of entertainment. You just have to assume that the plane is going to be destroyed on its first flight. Anything after that is bonus.
It's called Google Street View.
1. Why are you reinventing the wheel?
2. What makes you think that anyone wants to look at this?
3. Why do you want to waste your time babysitting a bunch of technology on vacation?
OH NOES!!!! SPOILED!!!! WE LIKE TOTALLLY NEVER SAW THOSE STILLS.
I can not jack my dick hard enough to a spoiler alerts.
SPOILER ALERT!!!!!!!!!
The Enterprise wins.
The "usually" here is a wonderful weasel word. You can't keep cutting taxes and seeing an increase in revenue. I mean think about it. If it was always true, why wouldn't you just cut your tax rate to zero and to maximize tax revenue? It doesn't make any sense. There's a point where you are simply cutting into revenue. Similarly, raising taxes will increase revenue, but there comes a point where they drag down the entire economy, thus cutting into revenues. There's a term for this, it's called "diminishing returns."
The fact is, that in the US the tax rate is one of the lowest of western world. In fact, according to the conservative American Enterprise Institute, tax revenue actually dropped as a result of the Bush tax cuts. The nonpartisan Congressional Budget Office came to the same conclusion in 2007.
Economies, especially modern economies, are complex interconnected systems. That means there are no simple solutions, and there are definitely no cure-alls. Those that constantly advocate tax cuts without regard to the situation or an understanding of their effects in the current economy, are simply advocating a means instead of ends. It's zealotry, and demonstrates no understanding of the problem.
I'll go with, "better than the tax cuts that decreased revenue."