[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [syndication] Re: RSS feed filtered by keywords?

Bill Kearney wrote:


Ok I at least see your point ... your operating on different assumption
than am I ... so at first I couldn't even see what your were saying.
I'm assuming that there are a group of feeds that you always read ...
your friends, your business associates, your personal network,  people
who have given you reliable or interesting information in the past.  You
will probably want to read every item in those feeds  .. or at least be
presented with every item.  But if you rely on that as your only source
of items, you will miss too much.   That is where keyword filtering
comes into play.  You apply the filters to the *entire river* of
information to discover new sources.


Well, you're halfway there.  I'm saying keyword filtering alone is a flawed
idea.  I suspect we agree in this regard.  I do recognize that keywords is where
we're going to have to start.  I just don't want to see them being viewed as
anything other than a bad hack.
I couldn't disagree more. Using keywords for discovery is what has gotten us to where we are today in the science and art of Information Retrieval. But I think there are two aspects to the use of keywords: 1) Discovering items out of the exact context of their authoring - in other words the author could not have anticipated the context of the reader. Using keywords is the only way I know to do that. 2) Authors (almost physically) placing items in a designated context - in other words the author assigns his item to a catagory. If the reader finds the context, they can find the items they want. I think this is the case that you are thinking about and are forgetting about case (1) above. This is where the author specifies catagories in their own feed - the author is then in total control of the context. 


Oh really ?   Can you read all the items in all the feeds in one day ?


Ah, do I need to?  Or do I take advantage of a larger network of what my peers
discover?  I'm telling you, this is going to be much more valuable than trying
to make each person 'hyper informed'.
Sometimes that works .... sometimes it don't ... depends on how you are focusing on a particular topic. 


If you store all the items from all the feeds that you have discovered
(regardless of whether those feeds have ever contained an item that
interests you) just how big and how fast will that database grow?  You
are one of the few people in the world that I think could actually give
us a reasonably accurate answer to that question.


Heh, yes the Syndic8 and other projects give me a pretty big picture.  There
are, at any given time, about 150k items in a day's feeds.  Something like 80mb,
IIRC.  It's not a massive amount.  It is just text after all and compresses to
considerably less.  But it /is/ a lot of items.  And yes, it is tough to "find"
the needle in the haystack.


Thanks, that's the answer to the question i asked here:
http://radio.weblogs.com/0113759/2003/01/09.html
... to which I never did get a good answer  ... till now :)




I agree, but that assumes that you have already built up such a network.
Many people have not, or are in the process of switching from one
network to another.  How are they to *rapidly* get oriented in a new
network ?


Whoa there, the point may not be to drag them into the mudpile and make them use
the existing networks.  The point may be to create new networks that don't
involve slogging through the mud.


... huh? .... could you elaborate? .... i have no idea what that means.



This why I came up with the idea of web-based shared subscription lists.  To be
able to use an online database (in this case Syndic8) to store the lists helps
in managing the massive amount of feeds.  You can then tell your reader program
to download from it.  That way more than one program can use the lists (home,
office, portable, etc).  These lists default to being private.  To get to the
next level it becomes worth sharing these lists.  That way you can see what
others are reading.  And as more people make their lists *programmatically*
accessible it'll be possible for software to start helping them see the bigger
pictures.


A great idea indeed !

Seth Russell
http://radio.weblogs.com/0113759/