I have written on my personal blog about what I am starting to call “Personal Reality“.

As I wrote there:

Personal Media includes your friend’s shared items. It includes the comments you leave on blogs. It includes Social Media. But it also includes private updates. Updates from your Intranet. Updates from your family. Updates from broadcast media. Updates that matter to you - no one else.

Personal Media is about recognizing that people are social and private. They are interested in personal experiences.

Check it out.

Why is this poll on CNN.com’s home page?

1. I don’t have a pet
2. I don’t want to own a pet
3. There are plenty of other more important things for CNN to cover than pets
4. Argg!

Come on people… please. How hard is it to learn my interests and serve up relevant content (I don’t even dare asking for APML support). Even without tracking user interests, I can almost guarantee you that people visiting CNN.com do not care about Dog food. Not on the front page!

An even broader question - do they not watch Jon Stewart? Do they not get it? The world is begging for real questions and real answers to real problems. How hard is it to stick to real news in this day and age. Surely they can leave Pet food to the Lifestyle channel?

Why do they waste our time with O.J Simpson? Ratings? Imagine the ratings they would get if they actually picked a fight with Washington - if they actually spelled out the truth of things for everyone to hear and see.

This is why Media 2.0 will win. We can use tools to find the real content and skip the garbage.

Showing pets love… buh.

According to the Wall Street Journal online Facebook is designing an ad system to use their extensive knowledge of its users to target advertising to them.

This move is hardly unexpected. Chances are many sites across many usage models are considering and implementing the same thing.

The WSJ writes:

Next year, Facebook hopes to expand on the service, one person says, using algorithms to learn how receptive a person might be to an ad based on readily available information about activities and interests of not just a user but also his friends — even if the user hasn’t explicitly expressed interest in a given topic. Facebook could then target ads accordingly.

The question, however, is how long users are going to accept having their information harvested and leveraged in this way - the very heart of the Attention Economy - without demanding portability and transparency.

The WSJ article continues:

While Facebook plans to protect its users’ privacy and possibly give them an option to keep certain information completely private, some Facebook users might rebel against the use of their personal information for the company’s gain.

And the perceptions that targeted ads create can be as much of a problem as the reality. “Most people don’t realize how targeting works; it becomes so good that even though it’s anonymous, you feel like they know you,” says Rishad Tobaccowala, CEO of Publicis Groupe-owned consulting firm Denuo Group. However, he says Facebook needs to be careful in implementing any targeted ad system, lest loyal users “find it creepy.”

This is key. Maintaining privacy is just a subset of giving users control. Control must include portability and transparency.

Using export/import formats like APML would soften the impact of privacy/control concerns. The problem is that walled gardens like Facebook (and yes - it is a walled garden) think that they need to lock users in in order to maintain their unique value.

The truth is, however, that unless Facebook begins to adopt more standards and open up its platform for export, it will be usurped by the first medium-scale network to do so. Don’t believe me? Remember that little network Facebook that blew Myspace & Linkedin up by opening up just a little?

Let’s hope that Facebook considers taking some measures before rolling out their new ad system.

Mark Lewis has written a piece over on Cnet about the need to flip the information delivery model. He writes:

“Web 2.0 flips the information delivery model upside down–it’s now about global access, and information at your fingertips, aggregated from sources that you don’t even necessarily know about, or care where they exist. Based on a set of search criteria, information in all its rich forms–media, video, audio, images, documents, text–all will be assembled together in context and delivered to users and applications for real-time experience.”

That’s a very poetic way of saying that in an age of hyper-choice, the most important challenge is to move beyond ‘What’s popular’ toward what’s ‘Personally Relevant‘.

I happen to also agree with Mark’s suggested implementation - Source agnostic aggregation filtered by persistent search (and Attention Profiling) and delivered in real-time.

We call it Particls.

With the announcement of Streamy and Thoof, however, Michael Arrington over on Techcrunch has declared that Personalized news is pointless and will never work.

He’s felt that way for a long time. I know… because he told me so while we were playing poker. A number of other people have suggested the same thing to me as well.

However, there are two things those people don’t understand.

1. Particls is not about Personalized News, it is about Personalized Alerting. We use the personalization part to rank content and determine how urgent the alert is for each user on an individualized basis.

   Thoof, Streamy and others are doing a very different (and worthwhile) job - and they are all potential partners of ours. We wish them the best of luck.

2. Just because something has not worked before does not mean it is not worth doing again and again until it’s done right. There is a place for popular, social news experiences (as Digg’s popularity has proved) and there is a place for targeted, personal and solitary news experiences (as Digg’s trolls and pop-culture content has proved).

Daniela recently pointed me to this Bear Stearns report via her blog post.

In it they make the same observations that I and others have been talking about for more than a year.

“User-Generated Content (UGC) Is Not a Fad…
Some investors remain skeptical that UGC is more than a passing fad. However, in our recent online video survey, UGC is the No. 1 and No. 2 most popular content category among men aged 18-34 (M18-34) and among all respondents, respectively. Moreover, if we define UGC as page views only from sites such as Myspace.com, Facebook.com, Youtube.com, Wikipedia.org, Blogger.com, and Digg.com (which is quite conservative), we estimate that UGC now accounts for 13% of total U.S. Internet traffic, up from 0%-1% in 2004. Based on these statistics, we submit that UGC is here to stay.”

Although using the term UGC is not great, their conclusion sounds very familiar to anyone reading this blog.

“apparent to us that as supply of video content rises, value will shift from content producers to aggregators and packagers of content that can best aid users in finding content that fits their specific interests”.

Of course, APML as a way of describing user interests, and Particls as a way of filtering and alerting users about new, personally relevant content, are both key technology pieces to this new media 2.0 reality.

Ian has a great write-up about why he loves APML.

He writes:

This got me thinking too, what if other more established places like Trustedplaces, Last.FM, etc also gave away a APML file as part of the profile of each user?

One of the things I loved about APML is the Implicit Data (U-AR) and Explicit Data (I-AM) elements. You can just imagine how simple it would be to output APML from something Last.FM. (whats below isn’t true APML markup, just my lazy json like writing)

Implicit (U-AR) last.fm {
concept{ Ferry Corsten = 0.87 }
concept{ Armin Van Buuren = 0.90 }
concept{ Sugar Babes = 0.1 }
concept{ Lemonhead = 0.00001 }
}

He also mentions being asked “What is the business imperative to support such a thing”.

In other words - if companies make their money from data lockin - then why would they want to give that data away.

I would suggest that anyone asking that question consider that publishers used to think like that. Now they all support RSS.

If feed readers thought like that, then OPML support and the rise of proper and continued innovation in the space may not have occurred.

If you are a smaller guy, supporting APML means that users can jump in and get started quickly. The barrier to entry is lowered.

If you are a bigger player it means an increasingly savvy user base will continue to trust your data mining activities. Also it means you can get a more complete picture of your users if they choose to share their APML from other services. It also means you become part of an ecosystem instead of a data silo - data silos are dead.

In the era of user empowerment, the business imperative is: play nice or users will move to other services that respect their rights. Just watch the mad rush to Facebook.

Jermiah, Web strategy guru at PodTech and fellow Media 2.0 Workgroup memeber has a great video interview with Eric Peterson posted on his blog.

They discuss Engagement, Attention an Analytics. Great chat - well worth a watch.

I have previously proposed some new concepts for syndicating Attention analytics information called Audient and Attent Streams which will go a long way in helping us move beyond pageviews and server logs.

I also appreciate Eric’s approach of multiple factors for measuring Engagement. Very much like our Personal Relevancy engine, we don’t think that any single factor or method is enough for determining the importance of an item to a user - so rather we create a complex algorithm that takes many things into account proportionally.

Check out the interview on Jeremiah’s Blog and check out Eric Peterson’s consulting company Web Analytics Demystified.

Every now and then someone asks ‘Why don’t you build Collaborative Recommendation into Particls’.

In case you don’t know, Collaborative Recommendation is when a system uses the recommendation of many people to ‘decide’ that a piece of content is worth seeing. So, like Digg for example, if 100 people vote that something is great (vote is a word I use loosely here), then it is probably worth seeing.

There are a few answers to that question.

1. Particls is really about filtering noise out - not discovering new recommended content (even though we provide some of that functionality just to get novice users started).
2. There are plenty of other great collaborative recommendation services out there, stick the RSS feed into Particls and there you go.
3. Google’s PageRank and Technorati’s Authority is already a form of Collaborative Recommendation - it isn’t really very new.
4. The next generation of Collaborative Recommendation is actually something different. Let’s call it Collaborative Filtering 3.0 or… Peer LifeStreams + Personal Relevancy

You have friends (hopefully); they have lifestreams (or at the very least RSS feeds from all their social/sharing sites) - plug their feeds into Particls and we filter out the stuff you don’t care about. What’s left? Stuff your friends ‘recommended’ that you actually care about. Collaborative Recommendation done right.

If you want to add me to your Collaborative Recommendation lineup, you can find me on Jaiku

Techcrunch reports that AOL will be upgrading their feed reader with new AJAX tricks and better OPML support. Check out the report on Techcrunch.

Also Nick of Feeddemon has released version 2.5 of his great desktop feed reader including item sharing through News Bins and a great popular topics feature.

It’s more than redundant to say now, but it’s clear that aggregation (in all forms) is here to stay and will be one of the primary ways people recieve and manage their information. From the mainstream AOL offerings to the power-user ready Feeddemon.

The next frontier, of course, is adding in Personal Relevancy.

Recently I have observed that we, as early adopters, use an enormous amount of implicit and explicit Metadata when making feed reading decisions.

When skimming our thousands of items a day we are actually making value judgements based on who the author is, what the headline reads (and what we think the topic is based on the headline), if there are any pictures to catch our eye and so on and so on.

When we come across a blog, I think that most people look for the subscriber count and consider (at least at the back of their mind and as part of a larger value judgement) whether or not they should add the author to their subscription list based on how authoritative that number makes them. Adding someone to your feed list is a relatively big decision. So the ’subscriber count’ metadata is important.

The problem though, is that mainstream users don’t know this metadata. They don’t know that engaget is the top gadget blog. They don’t know that Chris Messina is an authority on OpenID and Microformats and they don’t know what constitutes a small or large subscriber count. They also don’t know about Technorati and therefore don’t know how to check a blogs rank before consider the weight to place on the post.

R. Todd Stephens writes an article asking us to imagine a world without metadata. It’s a fascinating prospect.

He gives the following real-world example:

“Now imagine walking into your local grocery store, and you notice all of the traditional taxonomies have been removed because product classifications are a form of metadata. The aisle signage has been removed. The only things you can see are the blank containers designed for the products themselves. Let’s suppose you need soup to go with Saturday’s dinner. You grab a can and begin to shake it in hopes that the weight and movement can provide you with some indication of the contents. Is it tomato soup or a can of beans? Perhaps it is a can of peaches or mixed vegetables. Or, maybe you’re an experienced shopper who can distinguish between soup and other products. Is it chicken noodle soup, vegetable soup or clam chowder?”

He also talks about metadata without context using foreign travellers as an example.

“My wife and I ran across this in the Atlanta airport a few months ago when traveling overseas. A woman standing outside the train car that moved travelers from the concourse to the travel gates was having a problem understanding the metadata information that was all around her. She asked us if we knew any Spanish, to which my wife replied, “Un poquito,” or just a little. She started to reel off sentence after sentence, trying to explain to us her issues. The best we could do was to hand her off to another couple that knew much more Spanish than we did. Here is the point: as a traveler, she was surrounded by all the information and metadata she needed to either get her luggage or head to the departure gate. She simply couldn’t understand the information she needed to take action.”

I think that mainstream users are just like foreign travellers. They lack the understanding to use all the metadata ques to filter information quickly in a flooded feed reader.

I think that if mainstream Media and business management want to reach their audience, then we need to give them a way of helping users get important content by making metadata

a. Easier to understand.
b. Collectively factored and contributory to a single Personal Relevancy rank.