[Sigia-l] GoogleNews

[Listera]

"Michael Kay" wrote:

> No algorithm could consistently pick the 3 most relevant stories to lead with.
> (Even human editors have trouble figuring this out.)

I use several news aggregators to track news/info/data in about three dozen
fields every hour or so. (I used to use my own home-cooked app, later
Plucky, now I use NetNewsWire, on OSX.) It's an incredibly efficient way to
automatically get headlines, scan and, with a single click, get to the
actual news item on the web. This covers an enormous range, which would be
physically impossible for me to do manually.

NetNewsWire may bring scripted news item parsing, like filters one sets up
for email. So you can tell the app to peek into the news and make judgments
based on rules and conditionals you set up, as well as act on the results
by, say, automatically emailing you, sending an SMS message, getting more
news/info based on received data, eliminating headlines as being
less-than-relevant, intra-associating news items, sub-topic grouping, etc.

Everyday, I read hundreds of news items in this app. If it could track,
parse and detect patterns in my news reading habits, imagine the
possibilities. This is not far fetched.

Now imagine this sort of large scale user activity sampling yielding very
specific patterns that can be used by apps to make better decisions: like
Google (hypothetically) parsing your queries over time to decipher your
"needs", in news.

We're at the very beginning of this process. I'm not sure where it will go.
But I sure wouldn't dismiss it by saying machines will never be able to make
"intelligent" or "interesting" choices for us.

Best,

Ziya