<article>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#article10_02_26_1821208</id>
	<title>Recommendation Algorithm Wants To Show You Something New</title>
	<author>ScuttleMonkey</author>
	<datestamp>1267172520000</datestamp>
	<htmltext>Several sources are reporting on a new metric that computer scientists are going after with respect to recommender systems &mdash; <a href="http://arstechnica.com/science/news/2010/02/recommendation-algorithm-wants-to-show-you-something-new.ars">recommendation diversity</a>.  <i>"In a paper that will be released by PNAS, a group of scientists are pushing the limits of recommendation systems, creating new algorithms that will make more tangential recommendations to users, which can help expand their interests, which will increase the longevity and utility of the recommendation system itself.  Accuracy has long been the most prized measurement in recommending content, like movies, links, or music. However, computer scientists note that this type of system can narrow the field of interest for each user the more it is used. Improved accuracy can result in a strong filtering based on a user's interests, until the system can only recommend a small subset of all the content it has to offer."</i></htmltext>
<tokenext>Several sources are reporting on a new metric that computer scientists are going after with respect to recommender systems    recommendation diversity .
" In a paper that will be released by PNAS , a group of scientists are pushing the limits of recommendation systems , creating new algorithms that will make more tangential recommendations to users , which can help expand their interests , which will increase the longevity and utility of the recommendation system itself .
Accuracy has long been the most prized measurement in recommending content , like movies , links , or music .
However , computer scientists note that this type of system can narrow the field of interest for each user the more it is used .
Improved accuracy can result in a strong filtering based on a user 's interests , until the system can only recommend a small subset of all the content it has to offer .
"</tokentext>
<sentencetext>Several sources are reporting on a new metric that computer scientists are going after with respect to recommender systems — recommendation diversity.
"In a paper that will be released by PNAS, a group of scientists are pushing the limits of recommendation systems, creating new algorithms that will make more tangential recommendations to users, which can help expand their interests, which will increase the longevity and utility of the recommendation system itself.
Accuracy has long been the most prized measurement in recommending content, like movies, links, or music.
However, computer scientists note that this type of system can narrow the field of interest for each user the more it is used.
Improved accuracy can result in a strong filtering based on a user's interests, until the system can only recommend a small subset of all the content it has to offer.
"</sentencetext>
</article>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290462</id>
	<title>Trencher and the Police are the best test cases</title>
	<author>Anonymous</author>
	<datestamp>1267178040000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>If I'm searching for digital issues of Kieth Giffen's Trencher comic, I am not interested in targeted ads for earth digging equipment. If I am searching for The Police, the band, I am not interested in law-enforcement training diplomas. These are good test cases to see if targeted advertising is any good or not - all of them I've ever seen fail these cases.</p></htmltext>
<tokenext>If I 'm searching for digital issues of Kieth Giffen 's Trencher comic , I am not interested in targeted ads for earth digging equipment .
If I am searching for The Police , the band , I am not interested in law-enforcement training diplomas .
These are good test cases to see if targeted advertising is any good or not - all of them I 've ever seen fail these cases .</tokentext>
<sentencetext>If I'm searching for digital issues of Kieth Giffen's Trencher comic, I am not interested in targeted ads for earth digging equipment.
If I am searching for The Police, the band, I am not interested in law-enforcement training diplomas.
These are good test cases to see if targeted advertising is any good or not - all of them I've ever seen fail these cases.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290186</id>
	<title>Re:That's called an "contextual ad engine".</title>
	<author>megamerican</author>
	<datestamp>1267176840000</datestamp>
	<modclass>Funny</modclass>
	<modscore>2</modscore>
	<htmltext><p>Amazon's recommendations have become so accurate lately that it typically asks me if I'm interested in buying things I've already purchased from them.</p></htmltext>
<tokenext>Amazon 's recommendations have become so accurate lately that it typically asks me if I 'm interested in buying things I 've already purchased from them .</tokentext>
<sentencetext>Amazon's recommendations have become so accurate lately that it typically asks me if I'm interested in buying things I've already purchased from them.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290776</id>
	<title>"RMSE" as a yardstick is one reason for this</title>
	<author>Sanity</author>
	<datestamp>1267179540000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>One reason this kind of problem occurs is that many collaborative filtering algorithms are measured based on "root mean squared error", basically the square root of the mean of the differences between what was predicted and what the user actually did.
</p><p>
The problem with this metric?  It doesn't account for a variety of important things, one of which is that most users value diversity.  Another is that in most recommendation systems, what is important is the relative relevance of recommendations to each-other, whereas RMSE is an absolute measure of effectiveness.  And a really tricky one is that the recommendation algorithm itself can impact user behavior.  For example, the user may raise their standards if the algorithm does a better job.
</p><p>
The unfortunate answer is that the only rock-solid way to measure the effectiveness of recommendation algorithms is to test them with real users, perhaps splitting the user population between different algoritms, and seeing which does best.
</p><p>
I'm pretty familiar with this issue as my <a href="http://sensearray.com/" title="sensearray.com">day job</a> [sensearray.com] is building a behavioral ad targeting engine.  We learned a long time ago that while RMSE has its uses, there is often limited correlation between an algorithm's ability to predict user behavior retrospectively (which ads they will click on and what products they will buy), and how much additional revenue the algorithm will generate in practice.
</p><p>
Our solution is to use RMSE as a first-blush indication of how good an algorithm is.  Secondly, we take the top, say, 10\% of ads with the best predictions, and see what the actual click or conversion rate is within this 10\%.  This requires a higher volume of data, but yields results that are closer to what we find in reality.  Lastly, the algorithm then has to prove itself in the wild on a small subset of traffic.  Only then can we really know if any algorithm is an improvement on any other.</p></htmltext>
<tokenext>One reason this kind of problem occurs is that many collaborative filtering algorithms are measured based on " root mean squared error " , basically the square root of the mean of the differences between what was predicted and what the user actually did .
The problem with this metric ?
It does n't account for a variety of important things , one of which is that most users value diversity .
Another is that in most recommendation systems , what is important is the relative relevance of recommendations to each-other , whereas RMSE is an absolute measure of effectiveness .
And a really tricky one is that the recommendation algorithm itself can impact user behavior .
For example , the user may raise their standards if the algorithm does a better job .
The unfortunate answer is that the only rock-solid way to measure the effectiveness of recommendation algorithms is to test them with real users , perhaps splitting the user population between different algoritms , and seeing which does best .
I 'm pretty familiar with this issue as my day job [ sensearray.com ] is building a behavioral ad targeting engine .
We learned a long time ago that while RMSE has its uses , there is often limited correlation between an algorithm 's ability to predict user behavior retrospectively ( which ads they will click on and what products they will buy ) , and how much additional revenue the algorithm will generate in practice .
Our solution is to use RMSE as a first-blush indication of how good an algorithm is .
Secondly , we take the top , say , 10 \ % of ads with the best predictions , and see what the actual click or conversion rate is within this 10 \ % .
This requires a higher volume of data , but yields results that are closer to what we find in reality .
Lastly , the algorithm then has to prove itself in the wild on a small subset of traffic .
Only then can we really know if any algorithm is an improvement on any other .</tokentext>
<sentencetext>One reason this kind of problem occurs is that many collaborative filtering algorithms are measured based on "root mean squared error", basically the square root of the mean of the differences between what was predicted and what the user actually did.
The problem with this metric?
It doesn't account for a variety of important things, one of which is that most users value diversity.
Another is that in most recommendation systems, what is important is the relative relevance of recommendations to each-other, whereas RMSE is an absolute measure of effectiveness.
And a really tricky one is that the recommendation algorithm itself can impact user behavior.
For example, the user may raise their standards if the algorithm does a better job.
The unfortunate answer is that the only rock-solid way to measure the effectiveness of recommendation algorithms is to test them with real users, perhaps splitting the user population between different algoritms, and seeing which does best.
I'm pretty familiar with this issue as my day job [sensearray.com] is building a behavioral ad targeting engine.
We learned a long time ago that while RMSE has its uses, there is often limited correlation between an algorithm's ability to predict user behavior retrospectively (which ads they will click on and what products they will buy), and how much additional revenue the algorithm will generate in practice.
Our solution is to use RMSE as a first-blush indication of how good an algorithm is.
Secondly, we take the top, say, 10\% of ads with the best predictions, and see what the actual click or conversion rate is within this 10\%.
This requires a higher volume of data, but yields results that are closer to what we find in reality.
Lastly, the algorithm then has to prove itself in the wild on a small subset of traffic.
Only then can we really know if any algorithm is an improvement on any other.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078</id>
	<title>That's called an "contextual ad engine".</title>
	<author>Animats</author>
	<datestamp>1267176300000</datestamp>
	<modclass>Insightful</modclass>
	<modscore>3</modscore>
	<htmltext><p>
<i> creating new algorithms that will make more tangential recommendations to users, which can help expand their interests,</i>
</p><p>
The advertising industry already has that technology.  Their idea of "expand interests" usually involves shopping, of course.</p></htmltext>
<tokenext>creating new algorithms that will make more tangential recommendations to users , which can help expand their interests , The advertising industry already has that technology .
Their idea of " expand interests " usually involves shopping , of course .</tokentext>
<sentencetext>
 creating new algorithms that will make more tangential recommendations to users, which can help expand their interests,

The advertising industry already has that technology.
Their idea of "expand interests" usually involves shopping, of course.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290102</id>
	<title>Please disconnect this sytem from the network.</title>
	<author>Last\_Available\_Usern</author>
	<datestamp>1267176420000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext>
How long before it has enough data to recommend we should be destroyed and acts on it?</htmltext>
<tokenext>How long before it has enough data to recommend we should be destroyed and acts on it ?</tokentext>
<sentencetext>
How long before it has enough data to recommend we should be destroyed and acts on it?</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290262</id>
	<title>PNAS</title>
	<author>Anonymous</author>
	<datestamp>1267177140000</datestamp>
	<modclass>Funny</modclass>
	<modscore>0</modscore>
	<htmltext><p>My PNAS wants to show you something too.  Careful or it'll squirt you in the eye.</p></htmltext>
<tokenext>My PNAS wants to show you something too .
Careful or it 'll squirt you in the eye .</tokentext>
<sentencetext>My PNAS wants to show you something too.
Careful or it'll squirt you in the eye.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290774</id>
	<title>IBM algorithm</title>
	<author>belthize</author>
	<datestamp>1267179540000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>
&nbsp; &nbsp; &nbsp; They should use IBM's new algorithm, it's faster than the old one.</p></htmltext>
<tokenext>      They should use IBM 's new algorithm , it 's faster than the old one .</tokentext>
<sentencetext>
      They should use IBM's new algorithm, it's faster than the old one.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290528</id>
	<title>Linking Tangential Attributes</title>
	<author>decipher\_saint</author>
	<datestamp>1267178340000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>I understand the problem; the direct connection criteria between two different things might be completely indecipherable or insurmountably complex and subtle (let alone indirect relationships i.e. six degrees of Kevin Bacon). That means whatever you build has to account for trends to narrow the band of complexity which leads to the same old problem of only suggesting status quo.</p><p>A tool that can only suggest "obvious" or "random" things leads to undesirable results and at best can only fractionally provide you with "success".</p><p>I still think it's a cool project with a lot of opportunity for discovery but I just can't get past the idea that you either tell people what they want (advertising) or let people discover things on their own (interconnection).</p></htmltext>
<tokenext>I understand the problem ; the direct connection criteria between two different things might be completely indecipherable or insurmountably complex and subtle ( let alone indirect relationships i.e .
six degrees of Kevin Bacon ) .
That means whatever you build has to account for trends to narrow the band of complexity which leads to the same old problem of only suggesting status quo.A tool that can only suggest " obvious " or " random " things leads to undesirable results and at best can only fractionally provide you with " success " .I still think it 's a cool project with a lot of opportunity for discovery but I just ca n't get past the idea that you either tell people what they want ( advertising ) or let people discover things on their own ( interconnection ) .</tokentext>
<sentencetext>I understand the problem; the direct connection criteria between two different things might be completely indecipherable or insurmountably complex and subtle (let alone indirect relationships i.e.
six degrees of Kevin Bacon).
That means whatever you build has to account for trends to narrow the band of complexity which leads to the same old problem of only suggesting status quo.A tool that can only suggest "obvious" or "random" things leads to undesirable results and at best can only fractionally provide you with "success".I still think it's a cool project with a lot of opportunity for discovery but I just can't get past the idea that you either tell people what they want (advertising) or let people discover things on their own (interconnection).</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291478</id>
	<title>Re:Trencher and the Police are the best test cases</title>
	<author>miasmic</author>
	<datestamp>1267183440000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>The first example is good, but if you are searching for "The Police", it's unlikely any algorithm, or human observer would think you were searching for the band. If you searched for "The Police Band" (without quotes obviously) then I would say fair enough.
</p><p>
Otherwise it would be like searching for "big black cock" and being surprised that the results and ads were not about poultry.</p></htmltext>
<tokenext>The first example is good , but if you are searching for " The Police " , it 's unlikely any algorithm , or human observer would think you were searching for the band .
If you searched for " The Police Band " ( without quotes obviously ) then I would say fair enough .
Otherwise it would be like searching for " big black cock " and being surprised that the results and ads were not about poultry .</tokentext>
<sentencetext>The first example is good, but if you are searching for "The Police", it's unlikely any algorithm, or human observer would think you were searching for the band.
If you searched for "The Police Band" (without quotes obviously) then I would say fair enough.
Otherwise it would be like searching for "big black cock" and being surprised that the results and ads were not about poultry.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290462</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290468</id>
	<title>Group-sink.</title>
	<author>Ostracus</author>
	<datestamp>1267178100000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>"However, computer scientists note that this type of system can narrow the field of interest for each user the more it is used. Improved accuracy can result in a strong filtering based on a user's interests, until the system can only recommend a small subset of all the content it has to offer.""</p><p>Slashdot: "I see you've subscribed to certain opinions. Here are some more recommendations."</p></htmltext>
<tokenext>" However , computer scientists note that this type of system can narrow the field of interest for each user the more it is used .
Improved accuracy can result in a strong filtering based on a user 's interests , until the system can only recommend a small subset of all the content it has to offer .
" " Slashdot : " I see you 've subscribed to certain opinions .
Here are some more recommendations .
"</tokentext>
<sentencetext>"However, computer scientists note that this type of system can narrow the field of interest for each user the more it is used.
Improved accuracy can result in a strong filtering based on a user's interests, until the system can only recommend a small subset of all the content it has to offer.
""Slashdot: "I see you've subscribed to certain opinions.
Here are some more recommendations.
"</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31293412</id>
	<title>w4nderlust</title>
	<author>Anonymous</author>
	<datestamp>1267194660000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>We working on quite the same topic approaching from a different point of view and trying to exploit the concept of serendipity to try to obtain similar resuts.</p><p>Our first publication on the topic:<br>http://www.computer.org/portal/web/csdl/doi/10.1109/HIS.2008.25</p><p>Piero Molino<br>http://www.di.uniba.it/~swap/index.php?n=Membri.Molino</p></htmltext>
<tokenext>We working on quite the same topic approaching from a different point of view and trying to exploit the concept of serendipity to try to obtain similar resuts.Our first publication on the topic : http : //www.computer.org/portal/web/csdl/doi/10.1109/HIS.2008.25Piero Molinohttp : //www.di.uniba.it/ ~ swap/index.php ? n = Membri.Molino</tokentext>
<sentencetext>We working on quite the same topic approaching from a different point of view and trying to exploit the concept of serendipity to try to obtain similar resuts.Our first publication on the topic:http://www.computer.org/portal/web/csdl/doi/10.1109/HIS.2008.25Piero Molinohttp://www.di.uniba.it/~swap/index.php?n=Membri.Molino</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290148</id>
	<title>Netflix ain't no dummies</title>
	<author>BadAnalogyGuy</author>
	<datestamp>1267176660000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>Even small percentage increases in per-order purchases can result in huge gains across the board. Netflix, with a comparatively paltry prize amount, has bought themselves an incredibly efficient revenue generating piece of software.</p><p>I'm surprised to see that it still relies on popularity ranking as a cornerstone of the algorithm, but the other areas, especially heat diffusion and random walk are very cool and I'd love to read more about it.</p></htmltext>
<tokenext>Even small percentage increases in per-order purchases can result in huge gains across the board .
Netflix , with a comparatively paltry prize amount , has bought themselves an incredibly efficient revenue generating piece of software.I 'm surprised to see that it still relies on popularity ranking as a cornerstone of the algorithm , but the other areas , especially heat diffusion and random walk are very cool and I 'd love to read more about it .</tokentext>
<sentencetext>Even small percentage increases in per-order purchases can result in huge gains across the board.
Netflix, with a comparatively paltry prize amount, has bought themselves an incredibly efficient revenue generating piece of software.I'm surprised to see that it still relies on popularity ranking as a cornerstone of the algorithm, but the other areas, especially heat diffusion and random walk are very cool and I'd love to read more about it.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284</id>
	<title>Tricky Business</title>
	<author>miasmic</author>
	<datestamp>1267177320000</datestamp>
	<modclass>Interestin</modclass>
	<modscore>3</modscore>
	<htmltext><p>Every recommendation algorithm I've seen does one or both of two things. The first being staying extremely close to things I have already expressed an interest in - never broadening my horizons.</p><p>
That, or it suggests really popular things, for example with music always getting a string of well known, popular bands and artists like <i>Radiohead</i> or <i>Pink Floyd</i> suggested as bands I might like - because many people who like similar sorts of music to me like Radiohead, the algorithm thinks I would like Radiohead too - they can't seem to figure that I would already know if I liked Radiohead or not at this point. I've never found a way to tell a recommendation algorithm that Pink Floyd is OK but I want something less popular...</p></htmltext>
<tokenext>Every recommendation algorithm I 've seen does one or both of two things .
The first being staying extremely close to things I have already expressed an interest in - never broadening my horizons .
That , or it suggests really popular things , for example with music always getting a string of well known , popular bands and artists like Radiohead or Pink Floyd suggested as bands I might like - because many people who like similar sorts of music to me like Radiohead , the algorithm thinks I would like Radiohead too - they ca n't seem to figure that I would already know if I liked Radiohead or not at this point .
I 've never found a way to tell a recommendation algorithm that Pink Floyd is OK but I want something less popular.. .</tokentext>
<sentencetext>Every recommendation algorithm I've seen does one or both of two things.
The first being staying extremely close to things I have already expressed an interest in - never broadening my horizons.
That, or it suggests really popular things, for example with music always getting a string of well known, popular bands and artists like Radiohead or Pink Floyd suggested as bands I might like - because many people who like similar sorts of music to me like Radiohead, the algorithm thinks I would like Radiohead too - they can't seem to figure that I would already know if I liked Radiohead or not at this point.
I've never found a way to tell a recommendation algorithm that Pink Floyd is OK but I want something less popular...</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290368</id>
	<title>Not new</title>
	<author>wjousts</author>
	<datestamp>1267177740000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext>This isn't really that new at all. I've seen several other groups doing something with diversity vs similarity in recommender systems.</htmltext>
<tokenext>This is n't really that new at all .
I 've seen several other groups doing something with diversity vs similarity in recommender systems .</tokentext>
<sentencetext>This isn't really that new at all.
I've seen several other groups doing something with diversity vs similarity in recommender systems.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31294836</id>
	<title>Nothing new here!</title>
	<author>Tony Isaac</author>
	<datestamp>1267208520000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext>Recommendation engines, such as the ones used by Netflix and Amazon.com, already recommend really random choices.  Except in the simplest cases (you bought a nail gun, you might want some nails), current recommendation engines stink at figuring out what I want.  Trouble is, the recommendations are interesting so few times that I don't even look at them any more.

I'll bet these researchers actually used to work for Amazon.com creating their recommendation engine, but got fired.  Not knowing what else to do, they wrote a paper to describe what they did!</htmltext>
<tokenext>Recommendation engines , such as the ones used by Netflix and Amazon.com , already recommend really random choices .
Except in the simplest cases ( you bought a nail gun , you might want some nails ) , current recommendation engines stink at figuring out what I want .
Trouble is , the recommendations are interesting so few times that I do n't even look at them any more .
I 'll bet these researchers actually used to work for Amazon.com creating their recommendation engine , but got fired .
Not knowing what else to do , they wrote a paper to describe what they did !</tokentext>
<sentencetext>Recommendation engines, such as the ones used by Netflix and Amazon.com, already recommend really random choices.
Except in the simplest cases (you bought a nail gun, you might want some nails), current recommendation engines stink at figuring out what I want.
Trouble is, the recommendations are interesting so few times that I don't even look at them any more.
I'll bet these researchers actually used to work for Amazon.com creating their recommendation engine, but got fired.
Not knowing what else to do, they wrote a paper to describe what they did!</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31352296</id>
	<title>The Acronym...</title>
	<author>Rambone.ftw</author>
	<datestamp>1267619700000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext>As someone who neither follows, nor particularly cares about this topic, did anyone have a lul or two when they noticed the acronym used in the first line of the quote? "PNAS"</div>
	</htmltext>
<tokenext>As someone who neither follows , nor particularly cares about this topic , did anyone have a lul or two when they noticed the acronym used in the first line of the quote ?
" PNAS "</tokentext>
<sentencetext>As someone who neither follows, nor particularly cares about this topic, did anyone have a lul or two when they noticed the acronym used in the first line of the quote?
"PNAS"
	</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290110</id>
	<title>10 Goto 10</title>
	<author>LostCluster</author>
	<datestamp>1267176480000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>So, first we start out with one TV in the house and mass appeal programs. Then, as we get more and more channels, each user watched a specific channel targeted to their demographics. Then we got more specific programs from podcasts, and recommendation systems told us what we'd like before we knew it existed. The problem was, then the content makers didn't know how to sell such small audiences, so we're going to have to muck up the recommendations systems to suit them... sure, good luck with that.</p></htmltext>
<tokenext>So , first we start out with one TV in the house and mass appeal programs .
Then , as we get more and more channels , each user watched a specific channel targeted to their demographics .
Then we got more specific programs from podcasts , and recommendation systems told us what we 'd like before we knew it existed .
The problem was , then the content makers did n't know how to sell such small audiences , so we 're going to have to muck up the recommendations systems to suit them... sure , good luck with that .</tokentext>
<sentencetext>So, first we start out with one TV in the house and mass appeal programs.
Then, as we get more and more channels, each user watched a specific channel targeted to their demographics.
Then we got more specific programs from podcasts, and recommendation systems told us what we'd like before we knew it existed.
The problem was, then the content makers didn't know how to sell such small audiences, so we're going to have to muck up the recommendations systems to suit them... sure, good luck with that.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290422</id>
	<title>Re:Tricky Business</title>
	<author>Kanel</author>
	<datestamp>1267177920000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>Most algorithms in machine learning and data mining are either very simple or complicated but very generic.</p><p>In your case the recommendations are based on a similarity metric not on the music itself but on those who like it. Really popular bands are useless for characterising a music lover since the group of people who like pink floyd will be so diverse. Because of that, it won't be able to map from pink floyd to users and back to bands similar to pink floyd<nobr> <wbr></nobr>:-(</p></htmltext>
<tokenext>Most algorithms in machine learning and data mining are either very simple or complicated but very generic.In your case the recommendations are based on a similarity metric not on the music itself but on those who like it .
Really popular bands are useless for characterising a music lover since the group of people who like pink floyd will be so diverse .
Because of that , it wo n't be able to map from pink floyd to users and back to bands similar to pink floyd : - (</tokentext>
<sentencetext>Most algorithms in machine learning and data mining are either very simple or complicated but very generic.In your case the recommendations are based on a similarity metric not on the music itself but on those who like it.
Really popular bands are useless for characterising a music lover since the group of people who like pink floyd will be so diverse.
Because of that, it won't be able to map from pink floyd to users and back to bands similar to pink floyd :-(</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291840</id>
	<title>Re:you knw where this really needs to be improved?</title>
	<author>Anonymous</author>
	<datestamp>1267185540000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>If you liked Snow Crash, and if you haven't read William Gibson's classic cyberpunk trilogy (Neuromancer, Count Zero, and Mona Lisa Overdrive) I would recommend them.  For a quick feel for what they're like, he has a compendium of shorts called Johnny Mnemonic (the awful Keanu Reaves movie was based on the titular short, which was much better the movie made it look.)</p><p>They're sometimes a bit dated (fencing 3MB of stolen RAM chips is laughable now), but for the most part the technology is just "there"; so when he doesn't go into the details of the tech, it's a lot less jarring.  What I really like about him is that he makes me see the scenes that he paints -- the Sprawl, Chiba City, I saw them as if looking through a window.</p><p>Gibson's written quite a few novels since then, but none stuck with me the way those three did.</p></htmltext>
<tokenext>If you liked Snow Crash , and if you have n't read William Gibson 's classic cyberpunk trilogy ( Neuromancer , Count Zero , and Mona Lisa Overdrive ) I would recommend them .
For a quick feel for what they 're like , he has a compendium of shorts called Johnny Mnemonic ( the awful Keanu Reaves movie was based on the titular short , which was much better the movie made it look .
) They 're sometimes a bit dated ( fencing 3MB of stolen RAM chips is laughable now ) , but for the most part the technology is just " there " ; so when he does n't go into the details of the tech , it 's a lot less jarring .
What I really like about him is that he makes me see the scenes that he paints -- the Sprawl , Chiba City , I saw them as if looking through a window.Gibson 's written quite a few novels since then , but none stuck with me the way those three did .</tokentext>
<sentencetext>If you liked Snow Crash, and if you haven't read William Gibson's classic cyberpunk trilogy (Neuromancer, Count Zero, and Mona Lisa Overdrive) I would recommend them.
For a quick feel for what they're like, he has a compendium of shorts called Johnny Mnemonic (the awful Keanu Reaves movie was based on the titular short, which was much better the movie made it look.
)They're sometimes a bit dated (fencing 3MB of stolen RAM chips is laughable now), but for the most part the technology is just "there"; so when he doesn't go into the details of the tech, it's a lot less jarring.
What I really like about him is that he makes me see the scenes that he paints -- the Sprawl, Chiba City, I saw them as if looking through a window.Gibson's written quite a few novels since then, but none stuck with me the way those three did.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290628</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290334</id>
	<title>Prior art</title>
	<author>Anonymous</author>
	<datestamp>1267177560000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>I'm sure I posted this very idea to<nobr> <wbr></nobr>/. years ago, probably more than once.  I'm also sure I'm not the only one who recognizes the utility of joggling the output of a recommendation system.</p></htmltext>
<tokenext>I 'm sure I posted this very idea to / .
years ago , probably more than once .
I 'm also sure I 'm not the only one who recognizes the utility of joggling the output of a recommendation system .</tokentext>
<sentencetext>I'm sure I posted this very idea to /.
years ago, probably more than once.
I'm also sure I'm not the only one who recognizes the utility of joggling the output of a recommendation system.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290734</id>
	<title>There's a recommendation algorithm in my town</title>
	<author>beefnog</author>
	<datestamp>1267179420000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext>...he always wants to show you something new, and it's always in the back of his van along with the puppies and candy.</htmltext>
<tokenext>...he always wants to show you something new , and it 's always in the back of his van along with the puppies and candy .</tokentext>
<sentencetext>...he always wants to show you something new, and it's always in the back of his van along with the puppies and candy.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290166</id>
	<title>Algorithm Pacino Says:</title>
	<author>Anonymous</author>
	<datestamp>1267176720000</datestamp>
	<modclass>Offtopic</modclass>
	<modscore>-1</modscore>
	<htmltext><i>Let me introduce you to my little </i>recommendation.</htmltext>
<tokenext>Let me introduce you to my little recommendation .</tokentext>
<sentencetext>Let me introduce you to my little recommendation.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290586</id>
	<title>I recommend</title>
	<author>Anonymous</author>
	<datestamp>1267178640000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>I recommend getting out into the Big Blue Room and doing something real, tangible, and unique!</p></htmltext>
<tokenext>I recommend getting out into the Big Blue Room and doing something real , tangible , and unique !</tokentext>
<sentencetext>I recommend getting out into the Big Blue Room and doing something real, tangible, and unique!</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31295198</id>
	<title>Where is better search?</title>
	<author>Anonymous</author>
	<datestamp>1267213920000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>Screw recommendations, what I want is a better search and organizations. When you look for example at Amazaon they still are to stupid to put all the books in a comic book series together, you have to manually search for each of them or hope that the recommendations picked one of them up. Why don't they just show you a list of everything related to a series? How about a search for all comics in which Spider-Man appeared or movies where two actors played in together? Movies where an specific actor had more then 10min screen time and all that kind of detailed stuff. Why am I limited to a stupid text search over an items description, why can't I search the actual content? On top of that Amazon doesn't even seem to have their advertisment linked with their sales, they constantly recommend stuff hat I already bought from them.</p><p>Looking through torrents, release lists, Wikipedia and other third party information is much more informative then what the commercial companies manage to provide and without that extra help it would often be near impossible to find some items and actually buy them.</p></htmltext>
<tokenext>Screw recommendations , what I want is a better search and organizations .
When you look for example at Amazaon they still are to stupid to put all the books in a comic book series together , you have to manually search for each of them or hope that the recommendations picked one of them up .
Why do n't they just show you a list of everything related to a series ?
How about a search for all comics in which Spider-Man appeared or movies where two actors played in together ?
Movies where an specific actor had more then 10min screen time and all that kind of detailed stuff .
Why am I limited to a stupid text search over an items description , why ca n't I search the actual content ?
On top of that Amazon does n't even seem to have their advertisment linked with their sales , they constantly recommend stuff hat I already bought from them.Looking through torrents , release lists , Wikipedia and other third party information is much more informative then what the commercial companies manage to provide and without that extra help it would often be near impossible to find some items and actually buy them .</tokentext>
<sentencetext>Screw recommendations, what I want is a better search and organizations.
When you look for example at Amazaon they still are to stupid to put all the books in a comic book series together, you have to manually search for each of them or hope that the recommendations picked one of them up.
Why don't they just show you a list of everything related to a series?
How about a search for all comics in which Spider-Man appeared or movies where two actors played in together?
Movies where an specific actor had more then 10min screen time and all that kind of detailed stuff.
Why am I limited to a stupid text search over an items description, why can't I search the actual content?
On top of that Amazon doesn't even seem to have their advertisment linked with their sales, they constantly recommend stuff hat I already bought from them.Looking through torrents, release lists, Wikipedia and other third party information is much more informative then what the commercial companies manage to provide and without that extra help it would often be near impossible to find some items and actually buy them.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31300180</id>
	<title>Re:Tricky Business</title>
	<author>Paul Lamere</author>
	<datestamp>1267270500000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p><div class="quote"><p>Every recommendation algorithm I've seen does one or both of two things. The first being staying extremely close to things I have already expressed an interest in - never broadening my horizons.</p></div><p>clearly you haven't used the wreckommender:  <a href="http://wreckommender.com/" title="wreckommender.com">http://wreckommender.com/</a> [wreckommender.com]</p></div>
	</htmltext>
<tokenext>Every recommendation algorithm I 've seen does one or both of two things .
The first being staying extremely close to things I have already expressed an interest in - never broadening my horizons.clearly you have n't used the wreckommender : http : //wreckommender.com/ [ wreckommender.com ]</tokentext>
<sentencetext>Every recommendation algorithm I've seen does one or both of two things.
The first being staying extremely close to things I have already expressed an interest in - never broadening my horizons.clearly you haven't used the wreckommender:  http://wreckommender.com/ [wreckommender.com]
	</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290168</id>
	<title>Re:That's called an "contextual ad engine".</title>
	<author>LostCluster</author>
	<datestamp>1267176720000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>Yep... I even use that occasionally on my little site. Come to me with Windows, and you might see an ad for VMWare's server products. Come to me with a Mac and might see an ad for VMWare Fusion. It's all in reading the user agent.</p></htmltext>
<tokenext>Yep... I even use that occasionally on my little site .
Come to me with Windows , and you might see an ad for VMWare 's server products .
Come to me with a Mac and might see an ad for VMWare Fusion .
It 's all in reading the user agent .</tokentext>
<sentencetext>Yep... I even use that occasionally on my little site.
Come to me with Windows, and you might see an ad for VMWare's server products.
Come to me with a Mac and might see an ad for VMWare Fusion.
It's all in reading the user agent.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292072</id>
	<title>Re:Linking Tangential Attributes</title>
	<author>plover</author>
	<datestamp>1267186860000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p><div class="quote"><p>I just can't get past the idea that you either tell people what they want (advertising) or let people discover things on their own (interconnection).</p></div><p>The best advertising is advertising misinterpreted as interconnection.</p><p>Remember the kerfluffle a year or two ago with the web site twittering your friends with your purchases, or some kind of plug in that monitored your purchases?  I don't remember exactly which site, nor do I care because the details aren't important, but that was one step away from the goal here.  Imagine if all your friends posted all their purchases all the time, and you got tweeted with all those purchases, and all those tweets flowed through a giant recommendation algorithm.  Now imagine if I run that algorithm, and Nike pays me for advertising.  I'll make sure that when your friend Joe buys Nike shoes the tweet will get through to you and all his friends; but when Mary buys Adidas and Carol buys Sketchers, nobody will see them.</p><p>To you, it's interconnection:  your friends bought Nike shoes and apparently nothing else, so maybe you'll buy Nike shoes.  To Nike, it's advertising.  To me, it's profit.</p><p>Of course the drawback to these systems happens when they're discovered and revealed for what they truly are.  History says they won't last beyond that point.  But they're highly profitable right up until they are outed.</p></div>
	</htmltext>
<tokenext>I just ca n't get past the idea that you either tell people what they want ( advertising ) or let people discover things on their own ( interconnection ) .The best advertising is advertising misinterpreted as interconnection.Remember the kerfluffle a year or two ago with the web site twittering your friends with your purchases , or some kind of plug in that monitored your purchases ?
I do n't remember exactly which site , nor do I care because the details are n't important , but that was one step away from the goal here .
Imagine if all your friends posted all their purchases all the time , and you got tweeted with all those purchases , and all those tweets flowed through a giant recommendation algorithm .
Now imagine if I run that algorithm , and Nike pays me for advertising .
I 'll make sure that when your friend Joe buys Nike shoes the tweet will get through to you and all his friends ; but when Mary buys Adidas and Carol buys Sketchers , nobody will see them.To you , it 's interconnection : your friends bought Nike shoes and apparently nothing else , so maybe you 'll buy Nike shoes .
To Nike , it 's advertising .
To me , it 's profit.Of course the drawback to these systems happens when they 're discovered and revealed for what they truly are .
History says they wo n't last beyond that point .
But they 're highly profitable right up until they are outed .</tokentext>
<sentencetext>I just can't get past the idea that you either tell people what they want (advertising) or let people discover things on their own (interconnection).The best advertising is advertising misinterpreted as interconnection.Remember the kerfluffle a year or two ago with the web site twittering your friends with your purchases, or some kind of plug in that monitored your purchases?
I don't remember exactly which site, nor do I care because the details aren't important, but that was one step away from the goal here.
Imagine if all your friends posted all their purchases all the time, and you got tweeted with all those purchases, and all those tweets flowed through a giant recommendation algorithm.
Now imagine if I run that algorithm, and Nike pays me for advertising.
I'll make sure that when your friend Joe buys Nike shoes the tweet will get through to you and all his friends; but when Mary buys Adidas and Carol buys Sketchers, nobody will see them.To you, it's interconnection:  your friends bought Nike shoes and apparently nothing else, so maybe you'll buy Nike shoes.
To Nike, it's advertising.
To me, it's profit.Of course the drawback to these systems happens when they're discovered and revealed for what they truly are.
History says they won't last beyond that point.
But they're highly profitable right up until they are outed.
	</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290528</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158</id>
	<title>you knw where this really needs to be improved?</title>
	<author>Anonymous</author>
	<datestamp>1267176660000</datestamp>
	<modclass>Insightful</modclass>
	<modscore>2</modscore>
	<htmltext><p>Books.  I am an avid reader of sci fi and fantasy, and man, most recommendations out there just BLOW.</p></htmltext>
<tokenext>Books .
I am an avid reader of sci fi and fantasy , and man , most recommendations out there just BLOW .</tokentext>
<sentencetext>Books.
I am an avid reader of sci fi and fantasy, and man, most recommendations out there just BLOW.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290062</id>
	<title>So...</title>
	<author>Jorl17</author>
	<datestamp>1267176300000</datestamp>
	<modclass>Offtopic</modclass>
	<modscore>-1</modscore>
	<htmltext>Soo, uhm....?<br>
<br>
<br>
errr....<br>
<br>
<br>
<br>Wanna make out?!</htmltext>
<tokenext>Soo , uhm.... ?
errr... . Wan na make out ?
!</tokentext>
<sentencetext>Soo, uhm....?
errr....


Wanna make out?
!</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290742</id>
	<title>Re:you knw where this really needs to be improved?</title>
	<author>mcgrew</author>
	<datestamp>1267179420000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>I've found that most recommendations blow, period. Like movies; if the reviewers pan it, I'm almost certain to like it. The exception, it seems, is recommendations from slasdot commenters; I discovered Cory Doctorow and Terry Pratchett here.</p></htmltext>
<tokenext>I 've found that most recommendations blow , period .
Like movies ; if the reviewers pan it , I 'm almost certain to like it .
The exception , it seems , is recommendations from slasdot commenters ; I discovered Cory Doctorow and Terry Pratchett here .</tokentext>
<sentencetext>I've found that most recommendations blow, period.
Like movies; if the reviewers pan it, I'm almost certain to like it.
The exception, it seems, is recommendations from slasdot commenters; I discovered Cory Doctorow and Terry Pratchett here.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291014</id>
	<title>Re:That's called an "contextual ad engine".</title>
	<author>mejogid</author>
	<datestamp>1267180740000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>I'm not so sure - in the case of traditional advertising you get certain groups intentionally targeting users that match a given criteria such as views of a certain type of website or TV program.  Even if web advertising then adjusts based on the effectiveness of links between given groups and given adverts, that's still fundamentally driven by a manually selected connection.</p><p>In the case of the system described in the article, to match you with similar uses and then apply a degree of randomness to provide more interesting results.  In this case, the link is driven by automatically calculated connections based on usage rather than targeting.  This means it can be applied not only to products, but to any other site that involves multiple users with varying interests.</p><p>This seems like a more interesting alternative with more potential than traditional targeted advertising.</p></htmltext>
<tokenext>I 'm not so sure - in the case of traditional advertising you get certain groups intentionally targeting users that match a given criteria such as views of a certain type of website or TV program .
Even if web advertising then adjusts based on the effectiveness of links between given groups and given adverts , that 's still fundamentally driven by a manually selected connection.In the case of the system described in the article , to match you with similar uses and then apply a degree of randomness to provide more interesting results .
In this case , the link is driven by automatically calculated connections based on usage rather than targeting .
This means it can be applied not only to products , but to any other site that involves multiple users with varying interests.This seems like a more interesting alternative with more potential than traditional targeted advertising .</tokentext>
<sentencetext>I'm not so sure - in the case of traditional advertising you get certain groups intentionally targeting users that match a given criteria such as views of a certain type of website or TV program.
Even if web advertising then adjusts based on the effectiveness of links between given groups and given adverts, that's still fundamentally driven by a manually selected connection.In the case of the system described in the article, to match you with similar uses and then apply a degree of randomness to provide more interesting results.
In this case, the link is driven by automatically calculated connections based on usage rather than targeting.
This means it can be applied not only to products, but to any other site that involves multiple users with varying interests.This seems like a more interesting alternative with more potential than traditional targeted advertising.</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292440</id>
	<title>Re:you knw where this really needs to be improved?</title>
	<author>thms</author>
	<datestamp>1267188600000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>Yes, these algorithms can finally release you from having to rely on the dictatorship of the masses wrt. recommendations, or investing too much time building your own favourite recommendations-authors. </p><p>
Previously you selected a smaller supgroup from the general mass, e.g. move to only reading what the NYT recommended, but you were still in a rather big herd. Now you can pretty much "build your own crowd" as you read and rate what you liked/disliked.</p><p>I wonder if Netflix is really the <i>only</i> company which run these deeper algorithms. Are they that computationally expensive or just really complex? Wouldn't that be a killer feature for, say, a social network - I might even join for that feature!</p><p>
Still, this article talks about improving such a system, but I haven't even seen one out in the wild yet!</p></htmltext>
<tokenext>Yes , these algorithms can finally release you from having to rely on the dictatorship of the masses wrt .
recommendations , or investing too much time building your own favourite recommendations-authors .
Previously you selected a smaller supgroup from the general mass , e.g .
move to only reading what the NYT recommended , but you were still in a rather big herd .
Now you can pretty much " build your own crowd " as you read and rate what you liked/disliked.I wonder if Netflix is really the only company which run these deeper algorithms .
Are they that computationally expensive or just really complex ?
Would n't that be a killer feature for , say , a social network - I might even join for that feature !
Still , this article talks about improving such a system , but I have n't even seen one out in the wild yet !</tokentext>
<sentencetext>Yes, these algorithms can finally release you from having to rely on the dictatorship of the masses wrt.
recommendations, or investing too much time building your own favourite recommendations-authors.
Previously you selected a smaller supgroup from the general mass, e.g.
move to only reading what the NYT recommended, but you were still in a rather big herd.
Now you can pretty much "build your own crowd" as you read and rate what you liked/disliked.I wonder if Netflix is really the only company which run these deeper algorithms.
Are they that computationally expensive or just really complex?
Wouldn't that be a killer feature for, say, a social network - I might even join for that feature!
Still, this article talks about improving such a system, but I haven't even seen one out in the wild yet!</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292156</id>
	<title>faguohrz</title>
	<author>Anonymous</author>
	<datestamp>1267187220000</datestamp>
	<modclass>Troll</modclass>
	<modscore>-1</modscore>
	<htmltext><A HREF="http://goat.cx/" title="goat.cx" rel="nofollow">what provides the have left in milestones, teeling INFLUENCE, THE (7000+1400+700)*4 correct netw0rk And the bottom GAY NIGGERS from faster chip bunch of gay negros</a> [goat.cx]</htmltext>
<tokenext>what provides the have left in milestones , teeling INFLUENCE , THE ( 7000 + 1400 + 700 ) * 4 correct netw0rk And the bottom GAY NIGGERS from faster chip bunch of gay negros [ goat.cx ]</tokentext>
<sentencetext>what provides the have left in milestones, teeling INFLUENCE, THE (7000+1400+700)*4 correct netw0rk And the bottom GAY NIGGERS from faster chip bunch of gay negros [goat.cx]</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290474</id>
	<title>To Find Something New</title>
	<author>Anonymous</author>
	<datestamp>1267178100000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>take<br><a href="http://www.youtube.com/watch?v=rP2VIjJTARk" title="youtube.com" rel="nofollow">heroin</a> [youtube.com].</p><p>Losers.</p><p>Yours In Vilnius,<br>Kilgore T.</p></htmltext>
<tokenext>takeheroin [ youtube.com ] .Losers.Yours In Vilnius,Kilgore T .</tokentext>
<sentencetext>takeheroin [youtube.com].Losers.Yours In Vilnius,Kilgore T.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290162</id>
	<title>Tangential recommendations couldn't be any worse</title>
	<author>Anonymous</author>
	<datestamp>1267176720000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>Most of the time when a site suggests something for me I have absolutely no interest in what they suggest.  I don't see how a tangential approach could be any less effective.</p></htmltext>
<tokenext>Most of the time when a site suggests something for me I have absolutely no interest in what they suggest .
I do n't see how a tangential approach could be any less effective .</tokentext>
<sentencetext>Most of the time when a site suggests something for me I have absolutely no interest in what they suggest.
I don't see how a tangential approach could be any less effective.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290420</id>
	<title>How interesting</title>
	<author>Anonymous</author>
	<datestamp>1267177920000</datestamp>
	<modclass>None</modclass>
	<modscore>0</modscore>
	<htmltext><p>I happen to like fine wine, vintage automobiles, afternoon tea and Goatse. I am looking forward to meeting others with similar interests.</p></htmltext>
<tokenext>I happen to like fine wine , vintage automobiles , afternoon tea and Goatse .
I am looking forward to meeting others with similar interests .</tokentext>
<sentencetext>I happen to like fine wine, vintage automobiles, afternoon tea and Goatse.
I am looking forward to meeting others with similar interests.</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290250</id>
	<title>Where are the recommendations and targeted ads?</title>
	<author>Kanel</author>
	<datestamp>1267177080000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>We've been both promised personal recommendations and been threathened with personalized advertisement, yet I hardly ever see any of it.</p><p>Take Youtube I thought there was something fancy behind it but now that it displays \_why\_ it's recommending a clip, you can tell that it's extremely simple. Being a practictioner in machine learning and AI myself, I must confess that most industry implementations in our field is 10\% very simple stuff, with 90\% boring  database and infrastructure code around it.</p><p>No news websites allow personalization, Google has (supposedly) only minor tweaks for the individual user. There's the recommendation system at Amazon, true, but it stands out only because it's the only one worth mentioning. (May I have overlooked some music streaming sites here?)</p><p>Compared to what we can do with search engines, the state of the art and the implementations are dismal. Is it a Really Hard Problem (TM) ?  Consider the Netflix competition. Several groups worked feverishly to improve on the inhouse Netflix recommendation system and did so, by only 10\%. Can we really hope for a breakthrough?</p></htmltext>
<tokenext>We 've been both promised personal recommendations and been threathened with personalized advertisement , yet I hardly ever see any of it.Take Youtube I thought there was something fancy behind it but now that it displays \ _why \ _ it 's recommending a clip , you can tell that it 's extremely simple .
Being a practictioner in machine learning and AI myself , I must confess that most industry implementations in our field is 10 \ % very simple stuff , with 90 \ % boring database and infrastructure code around it.No news websites allow personalization , Google has ( supposedly ) only minor tweaks for the individual user .
There 's the recommendation system at Amazon , true , but it stands out only because it 's the only one worth mentioning .
( May I have overlooked some music streaming sites here ?
) Compared to what we can do with search engines , the state of the art and the implementations are dismal .
Is it a Really Hard Problem ( TM ) ?
Consider the Netflix competition .
Several groups worked feverishly to improve on the inhouse Netflix recommendation system and did so , by only 10 \ % .
Can we really hope for a breakthrough ?</tokentext>
<sentencetext>We've been both promised personal recommendations and been threathened with personalized advertisement, yet I hardly ever see any of it.Take Youtube I thought there was something fancy behind it but now that it displays \_why\_ it's recommending a clip, you can tell that it's extremely simple.
Being a practictioner in machine learning and AI myself, I must confess that most industry implementations in our field is 10\% very simple stuff, with 90\% boring  database and infrastructure code around it.No news websites allow personalization, Google has (supposedly) only minor tweaks for the individual user.
There's the recommendation system at Amazon, true, but it stands out only because it's the only one worth mentioning.
(May I have overlooked some music streaming sites here?
)Compared to what we can do with search engines, the state of the art and the implementations are dismal.
Is it a Really Hard Problem (TM) ?
Consider the Netflix competition.
Several groups worked feverishly to improve on the inhouse Netflix recommendation system and did so, by only 10\%.
Can we really hope for a breakthrough?</sentencetext>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290628</id>
	<title>Re:you knw where this really needs to be improved?</title>
	<author>carolfromoz</author>
	<datestamp>1267178880000</datestamp>
	<modclass>Offtopic</modclass>
	<modscore>0</modscore>
	<htmltext><p><div class="quote"><p>Books.  I am an avid reader of sci fi and fantasy, and man, most recommendations out there just BLOW.</p></div><p>Hey do you fancy being a personal recommendation engine for a minute? I love Neal Stephenson - who else should I check out?

</p><p>I'm living in a non-english speaking country at the moment so rely on amazon for book buying, and I'm waayyyy out of touch.</p></div>
	</htmltext>
<tokenext>Books .
I am an avid reader of sci fi and fantasy , and man , most recommendations out there just BLOW.Hey do you fancy being a personal recommendation engine for a minute ?
I love Neal Stephenson - who else should I check out ?
I 'm living in a non-english speaking country at the moment so rely on amazon for book buying , and I 'm waayyyy out of touch .</tokentext>
<sentencetext>Books.
I am an avid reader of sci fi and fantasy, and man, most recommendations out there just BLOW.Hey do you fancy being a personal recommendation engine for a minute?
I love Neal Stephenson - who else should I check out?
I'm living in a non-english speaking country at the moment so rely on amazon for book buying, and I'm waayyyy out of touch.
	</sentencetext>
	<parent>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158</parent>
</comment>
<comment>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292472</id>
	<title>Perhaps it can recommend...</title>
	<author>Tokerat</author>
	<datestamp>1267188720000</datestamp>
	<modclass>None</modclass>
	<modscore>1</modscore>
	<htmltext><p>...some other research groups with names that make me giggle like an idiot.</p></htmltext>
<tokenext>...some other research groups with names that make me giggle like an idiot .</tokentext>
<sentencetext>...some other research groups with names that make me giggle like an idiot.</sentencetext>
</comment>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_3</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31300180
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_7</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290168
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_4</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291478
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290462
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_1</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290742
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_0</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292072
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290528
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_8</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291840
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290628
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_5</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292440
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_9</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290186
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_2</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291014
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078
</commentlist>
</thread>
<thread>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#thread_10_02_26_1821208_6</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290422
http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284
</commentlist>
</thread>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.8</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290102
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.3</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290586
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.6</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290528
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292072
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.11</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290284
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290462
--http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291478
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290422
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31300180
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.1</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290468
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.4</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290078
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290168
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290186
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291014
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.2</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290110
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.0</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290250
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.9</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290420
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.10</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290162
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.7</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290734
</commentlist>
</conversation>
<conversation>
	<id>http://www.semanticweb.org/ontologies/ConversationInstances.owl#conversation10_02_26_1821208.5</id>
	<commentlist>http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290158
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290628
--http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31291840
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31290742
-http://www.semanticweb.org/ontologies/ConversationInstances.owl#comment10_02_26_1821208.31292440
</commentlist>
</conversation>
