I'm also trying to get into this APML thing. But I have some doubts if its benefits really outweigh all privacy issues I see there.
David Novakovic
· 2 years ago
Theoretically, the user has control of their attention profile. Combined with using services that respect your privacy, and being able to remove things that you don't want in your profile, I think it will be less of an issue.
Jason Grunstra
· 2 years ago
From a privacy standpoint it seems like you could easily tap into a social bookmarking service like del.icio.us where if you add something there it is automatically added to your APML.
If that tag cloud could then programmatically follow you around the web and watch your browsing behavior -- to only intelligently add new sites/content to your APML based on cloud matches -- it should work pretty well in filtering out any "porn" or "militant Islamic websites" since most likely you wont be adding those types of tags to your social bookmarks.
Chris Saad
· 2 years ago
I agree Jason - APML does not need to be more complex that converting your Del.icio.us bookmarks (which are already public) to a portable list of interests.
It is up to the user which tools they use, which APML they export and who they give it to.
The only goal here is to standardize the format so the user has the choice.
I convert my Jaiku lifestream to APML for example - all that stuff is *already* public and manually selected by me.
Nicholas Carlson
· 2 years ago
Sorry, but APML, schmay-PML. It's all about Valleywag Lab's all new patented “Fanafi†algorithm — that’s short for “Find A Need And Fill It,†folks. It uses machine-learning semantic-graph discovery technologies to extract the needs users express when they add a Facebook app. It's going to get developers rich, quick. Come check it out.
Mark "Rizzn" hopkins
· 2 years ago
@Jason Grunstra: I read somewhere in my research that one of the needs for APML was increased user control of what's in your profile, essentially, because groups like Google are monitoring your browsing history ad infinitum. It was suggested that either through browser plugin or Google becoming APML aware that APML also be generated off browsing history. This is where the root of my concerns arose.
Mark "Rizzn" Hopkins
· 2 years ago
@Chris Saad: First, thanks for your help in putting this journey together for me. But for you and Marj, I would have been completely negative on APML.
I still have reservations about APML, but as I discussed on my podcast today with my cohost, there will still be a strong contingent of folks that feel that APML and the contents there-of are absolutely no-one's business, and will not comply at any nudging or cajoling.
Pete
· 2 years ago
Use the "reply" button - easier that "@". :P
Awesome article, by the way.
Mark "Rizzn" Hopkins
· 2 years ago
Hah - thanks. Adam just pointed that out to me too.
Jonnyajax
· 2 years ago
Great post Mark.
At Cluztr, we've been supporting APML created from each users clickstream - which they control and manage. We've gotten great feedback so far the output is an increasingly accurate picture of one's interests.
Amazon, Google and all kinds of other services have been leveraging this kind of attention profiling for years. It's about the time the user can start using it for their own benefit.
Privacy issues are marginal compared to the benefits and potential uses of APML.
Besides, the kids today don't care about privacy.
michael j pastor
· 2 years ago
jonnyajax, allow me to rephrase your comment to a different concept
...Besides, with treatment options as advanced as they are, kids today don't care about having safe sex either, but that doesn't mean AIDS has gone away.
Privacy issues are never a marginal issue, and thank you for alerting me to your company's practices in regards to clickstreams and your attitude about your user's privacy. I will be sure *not* to use cluztr from this point on.
One does not need to reveal anything about one's self to utilize the concept of an APML construct. Sorting and ranking of interests is a basic psychological construct, and it's something that we do *internally*. I don't need to tell a web service *all* of my interests to tell them what I'm interested in getting from them. There's no need for portability with a concept that never moves from where it belongs - with me, on my computer, in my head.
We don't announce *all* of our interests to the outside world when engaging with a particular person about a particular topic. NOT collecting information about pron tells me just as much about you as actually doing it (and in some ways, more - either you are a sexually repressed neurotic robot or you're blatantly lying and using another browser to look at pron).
It is *not* as simple as turning your tag cloud from delicious into something portable, because not all tags are created equal. Sorting and ranking is a highly complex internal judgment process that compares N^N concepts to each other.
APML is yet another attempt to create the One-Ring-to-Rule-Them-All (like OpenID). You don't need One Ring, you need Many-Specific-and-Appropriate Rings. But most importantly, you need an application that *MANAGES* all of those rings, and nobody else needs to see what you do within that application, and that application doesn't need to reside anywhere except on your desktop computer of choice. Outside services just need to given the appropriate Ring (let's be cute and call it a Friendship Ring), not my entire collection of Rings.
The whole point of the hypertext and the internet is that information is distributed, so why do all of these companies keep trying to consolidate everything?
The death of Web 2.0 will be when people realize that they don't need to do everything publicly and blatantly to get what they want, and you don't need a company that drops the "e" in -er to do it.
Jonnyajax
· 2 years ago
Michael, You're obviously very pasionate about your privacy, as are others who use social apps, but not everyone.
The explosive popularity of online services like Last.fm, iLike and Pandora have shown that people are more willing than ever to give up aspects of their privacy in exchange for the benefits of social networking and discovery.
To explain my statement that "privacy issues are marginal ..." was with the assumption that whatever social service has the appropriate privacy features in place for their users, whether the user chooses to leverage the option or not.
re: "thank you alerting me to your company's practices in regards to clickstreams and your attitude about your user's privacy." I suggest you read our Privacy Policy and TOS. Cluztr follows the guidelines set out by AttentionTrust.org, giving users full control, ownership and transparency in all aspects of how their data is handled.
My personaly attitude about a user's privacy is that it is inherent to the user in question, therefore, we provide numerous privacy features to accomodate everyone.
michael j pastor
· 2 years ago
I think the popularity of *all* present social networking sites is merely because *there has not been any kind of personal alternative of ANY kind*
You don't need to give up ANY of your privacy rights in ANY way to gain the benefits of social networking, and I'm personally (especially on the eve of Google's announcement of OpenSocial) completely flummuxed as to why nobody understands this.
I don't need to upload my address books to others' websites to network - my address book IS the network. All giving up any information to any site does is empower the company to make money from my information, when the real person who should be making money from my information is ME, the end user.
It's impossible for a social networking site to provide privacy practices because it is INHERENTLY IMPOSSIBLE for them to do it because the very premise of social networking sites as they are imagined in the current day is PUBLIC.
WHERE are the private desktop versions of all of this web 2.0 crap? Nowhere, because the plethora of beta-driven websites on steroids couldn't survive on a desktop, and most of all, they can't find a way to monetize it.
Elliot
· 2 years ago
Great post, lots of buzz about APML in recent weeks.
The summarization of potential stumbling blocks was interesting, but I have to disagree with the basis behind the "Not Everybody Reads the News" argument.
This assumes that APML is only applicable to news-reader applications, RSS/ATOM in general, or those that frequent news websites. This is definitely not the case. APML is a generic standard that can be utilized to express attention preferences in a variety of settings.
As more companies being embracing APML, we'll see a lot of interesting applications come out of the woodwork. At Orchestr8, we're integrating an APML engine into our product that enables users to "implicitly" update their interests based on web browsing activity, and leverage APML data to perform "conditional" operations (like highlighting webpage text or emailing content based on your attention preferences). These examples are web-focused, but there's nothing preventing APML from being used in other contexts, as well (IM, Email, etc.).
That said, there are potential issues moving forward. These include the Privacy concerns expressed in this post, synchronization / portability issues, and other concerns. However, APML certainly represents an exciting development in the world of 'structured attention'. We're excited to be supporting APML in the next release of our AlchemyPoint software.
James Gillmore
· 2 years ago
Won't Open Social be able to do exactly this, but better?
Mark "Rizzn" Hopkins
· 2 years ago
I'm not sure I'm the best person to ask about this... I think OpenSocial has some capabilities in this direction, but not quite all inclusive enough to qualify to serve the purposes of APML.
I've not completely gone over all the public specs of OpenSocial as muc h as I'd like... so I'm not sure if my answer is entirely accurate.
Wayne Smallman
· 1 year ago
Excellent article, and very much like follow-up material to my: "Will advertising ever not be annoying" article from earlier this year...
Shaiya
· 1 year ago
When I played the game for the first time ,I liked it deeply until present.In this game ,we can use the shaiya money and the shaiya gold .These shaiya money that if we got a upgrading ,maybe we could get a encourage for these shaiya money .But it is not the only way to get these shaiya money,we can also use the RMB to buy shaiya gold.
james
· 1 year ago
thiemehennis
· 11 months ago
So what happened to Cluztr and Particls? Have been looking at their websites/blogs/twitter accounts and no activity.. strange for an "activity profiling" website. BTW. I believe in APML-ish standards becoming very important, but it seems that the initiatives mentioned in this discussion have not really been making a lot of progress. Also, the APML wiki and discussion groups are not very active.. what happened?
miracLes
· 5 months ago
Thank you very much for this information. Good post thanks for sharing.
All Comments
I'm also trying to get into this APML thing. But I have some doubts if its benefits really outweigh all privacy issues I see there.
If that tag cloud could then programmatically follow you around the web and watch your browsing behavior -- to only intelligently add new sites/content to your APML based on cloud matches -- it should work pretty well in filtering out any "porn" or "militant Islamic websites" since most likely you wont be adding those types of tags to your social bookmarks.
It is up to the user which tools they use, which APML they export and who they give it to.
The only goal here is to standardize the format so the user has the choice.
I convert my Jaiku lifestream to APML for example - all that stuff is *already* public and manually selected by me.
I still have reservations about APML, but as I discussed on my podcast today with my cohost, there will still be a strong contingent of folks that feel that APML and the contents there-of are absolutely no-one's business, and will not comply at any nudging or cajoling.
Awesome article, by the way.
At Cluztr, we've been supporting APML created from each users clickstream - which they control and manage. We've gotten great feedback so far the output is an increasingly accurate picture of one's interests.
Amazon, Google and all kinds of other services have been leveraging this kind of attention profiling for years. It's about the time the user can start using it for their own benefit.
Privacy issues are marginal compared to the benefits and potential uses of APML.
Besides, the kids today don't care about privacy.
...Besides, with treatment options as advanced as they are, kids today don't care about having safe sex either, but that doesn't mean AIDS has gone away.
Privacy issues are never a marginal issue, and thank you for alerting me to your company's practices in regards to clickstreams and your attitude about your user's privacy. I will be sure *not* to use cluztr from this point on.
One does not need to reveal anything about one's self to utilize the concept of an APML construct. Sorting and ranking of interests is a basic psychological construct, and it's something that we do *internally*. I don't need to tell a web service *all* of my interests to tell them what I'm interested in getting from them. There's no need for portability with a concept that never moves from where it belongs - with me, on my computer, in my head.
We don't announce *all* of our interests to the outside world when engaging with a particular person about a particular topic. NOT collecting information about pron tells me just as much about you as actually doing it (and in some ways, more - either you are a sexually repressed neurotic robot or you're blatantly lying and using another browser to look at pron).
It is *not* as simple as turning your tag cloud from delicious into something portable, because not all tags are created equal. Sorting and ranking is a highly complex internal judgment process that compares N^N concepts to each other.
APML is yet another attempt to create the One-Ring-to-Rule-Them-All (like OpenID). You don't need One Ring, you need Many-Specific-and-Appropriate Rings. But most importantly, you need an application that *MANAGES* all of those rings, and nobody else needs to see what you do within that application, and that application doesn't need to reside anywhere except on your desktop computer of choice. Outside services just need to given the appropriate Ring (let's be cute and call it a Friendship Ring), not my entire collection of Rings.
The whole point of the hypertext and the internet is that information is distributed, so why do all of these companies keep trying to consolidate everything?
The death of Web 2.0 will be when people realize that they don't need to do everything publicly and blatantly to get what they want, and you don't need a company that drops the "e" in -er to do it.
You're obviously very pasionate about your privacy, as are others who use social apps, but not everyone.
The explosive popularity of online services like Last.fm, iLike and Pandora have shown that people are more willing than ever to give up aspects of their privacy in exchange for the benefits of social networking and discovery.
To explain my statement that "privacy issues are marginal ..." was with the assumption that whatever social service has the appropriate privacy features in place for their users, whether the user chooses to leverage the option or not.
re: "thank you alerting me to your company's practices in regards to clickstreams and your attitude about your user's privacy."
I suggest you read our Privacy Policy and TOS. Cluztr follows the guidelines set out by AttentionTrust.org, giving users full control, ownership and transparency in all aspects of how their data is handled.
My personaly attitude about a user's privacy is that it is inherent to the user in question, therefore, we provide numerous privacy features to accomodate everyone.
You don't need to give up ANY of your privacy rights in ANY way to gain the benefits of social networking, and I'm personally (especially on the eve of Google's announcement of OpenSocial) completely flummuxed as to why nobody understands this.
I don't need to upload my address books to others' websites to network - my address book IS the network. All giving up any information to any site does is empower the company to make money from my information, when the real person who should be making money from my information is ME, the end user.
It's impossible for a social networking site to provide privacy practices because it is INHERENTLY IMPOSSIBLE for them to do it because the very premise of social networking sites as they are imagined in the current day is PUBLIC.
WHERE are the private desktop versions of all of this web 2.0 crap? Nowhere, because the plethora of beta-driven websites on steroids couldn't survive on a desktop, and most of all, they can't find a way to monetize it.
The summarization of potential stumbling blocks was interesting, but I have to disagree with the basis behind the "Not Everybody Reads the News" argument.
This assumes that APML is only applicable to news-reader applications, RSS/ATOM in general, or those that frequent news websites. This is definitely not the case. APML is a generic standard that can be utilized to express attention preferences in a variety of settings.
As more companies being embracing APML, we'll see a lot of interesting applications come out of the woodwork. At Orchestr8, we're integrating an APML engine into our product that enables users to "implicitly" update their interests based on web browsing activity, and leverage APML data to perform "conditional" operations (like highlighting webpage text or emailing content based on your attention preferences). These examples are web-focused, but there's nothing preventing APML from being used in other contexts, as well (IM, Email, etc.).
That said, there are potential issues moving forward. These include the Privacy concerns expressed in this post, synchronization / portability issues, and other concerns. However, APML certainly represents an exciting development in the world of 'structured attention'. We're excited to be supporting APML in the next release of our AlchemyPoint software.
I've not completely gone over all the public specs of OpenSocial as muc h as I'd like... so I'm not sure if my answer is entirely accurate.
Good post thanks for sharing.
fx15 lida yılan yağı karınca yumurtası xacc
dizin