Smartening the aggregator

Steve Borche of Connecting the Dots has an idea for "smart aggregators" in Information Overload: Can you see what's coming?  This is something I've been looking for as well for some of the same reasons, primarily information overload.  But also for better flow (dare I say, aggregation) of the river that flows through my aggregator.

[The] content deluge -- and identifying trends and the nuggets of useful information -- is a huge problem for me and for you regardless of what industry you're in or job you have. It will become an even bigger issue when content producers around the world are increasingly accessible (with automatic translation getting more robust and accurate) and people interested in a topic or category are then compelled to *widen* the scope of information they want to consume, analyze and understand.

[via Alex Barnett, smart aggregators]

Borche makes the distinction between socially smart aggregation and smart aggregation that is focused on me (the user).  Socially smart aggregators are tools that monitor a given "social circle" and highlight websites / articles that appear to be interesting to the community, based on collective action within that community.  If you're a big technology person, then Tech Memeorandum is for you.  Borche says that he doesn't necessarily want everything that Tech Memeorandum picks up -- he wants this to allow for more facets, so that he can filter based on his particular interests or his particular circle of people.  The next step will be tools that let me filter based on a pre-specified list of others (and David Weinberger is asking for current examples).  He wants a smart aggregator.

Smart aggregators let me do some of the socially smart stuff, such as further filtering of Tech Memeorandum.  But it lets me mix up my own feeds and deal with them as I wish.  Within my own aggregator, I would like to

  • Highlight a post and see everyone that is talking about it in my current feeds, and/or on the web at large.
  • Highlight a given topic and see all the articles that reference it quickly within my feeds (and beyond).
  • Make these filters persistent, so the next time the aggregator gets news, I can quickly see the new "interesting" stuff.
  • Figure out a way to train the filters to help me find new stuff that I might think is interesting, like the Bayesian spam filters.

As I've mentioned before, I like threading as implemented in SharpReader.  I also like the direction BlogBridge and GreatNews are going with their persistent filters.  I believe GreatNews let me set a filter that always marks items as read, such as "topic x" that I never want to read if it appears in a post title.  Can someone bring all these things together?

Some of my previous thoughts on aggregator features:

6 Comment(s)

Items 1-3 of your short list above sound good, but I am very protective of my serendipitous discovery. I don't want things brought to the top of my feeds because they are similar to what I've clicked through before, for example. I've found that organizing feeds is themed folders works well, and clicking on "river of news" when I have the time usually shows me what the hot topics are because they get enough buzz to appear in more than one feed I scan. Perhaps some "waypath" -like option would be nice to be able to click on, but (I can't believe I'm saying this) I'd rather see RSS stay the way it is than take too big a risk on seeing it ruined.

jackvinson Author Profile Page said:

We are in agreement, Marshall. One of the problems I have with services that explicitly monitor everything I click / view is that I don't think that would give a very good perspective on what I really want. I already group my feeds by folder, so I can focus on general topics when necessary.

What I was thinking with the trainable filter option was that I would do the training by saying, "add articles like this" to a filter. I supose it would have to have a "but not like this" option as well.

I don't want my feeds gunked up with extraneous stuff either. I think it's my aggregator that should give me options to find similar / linked articles within my feeds or on the larger web. SharpReader has an option to see the Technorati Cosmos on an article, but they are usually too fresh for that to be of much help. I'd like to see more of those kinds of options built into the aggregator. Maybe there could be a standard interface that lets users search on their favorite blog search tool.

Here here on all that Jack. Ever notice how slow all the blogsearch engines are to discover inbound links anyway? I just got notified by Technorati of a link to one of my posts that I've been commenting back on and seeing loads of traffic from for like 48 hours - it's almost always my Statcounter that shows me inbound links before any of the blogsearch engines do. I would like to see which one is quickest, and I'm guessing it's going to be Google Blogsearch. So attach that to my feeds and I'd be happiest, I'm guessing.

» Managing Information Overload from New Persuasion

On Newsgator, I subscribe to 370 different feeds - right now there are 4765 posts from these feeds waiting for me to read them. Sometimes I give up, read my favorites, erase the rest and start from scratch - but Read More

» Text mining for the rest of us from Knowledge Jolt with Jack

Fellow Corante Web Hub contributor, Matthew Hurst wants Consumer Facing Text Mining Opportunities. He wants text mining for non-commercial customers that actually provides some value to the user. Read More

Jack Pan said:

Jack, I found your site through referrers to GreatNews. I must say your "Ideal feed reader features" post is one of the most insightful thoughts about rss reader usability. Subscribed!

Leave a comment


Previous entry: Communities and the web

Next entry: Human elements in the KM puzzle

Picture a steaming coffee cup. Better yet, grab one and have a read!

KJolt Memberships

Follow jackvinson on Twitter

View Jack Vinson's profile on LinkedIn