Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Google Living Stories (googlelabs.com)
16 points by physcab on Dec 8, 2009 | hide | past | favorite | 12 comments


Great, they're going the exact opposite direction of what I'm willing to pay for.

I wish my local paper(s) would cancel their Monday through Saturday editions and concentrated on one great Sunday paper. I'd pay ten times as much for clear, concise, complete "dead" stories on dead trees.


Exactly - tell me all the local stuff without all the AP global/national stuff in between. I would absolutely subscribe to this and would even prefer it to an online version.


Looking at this made me realize that the only news sites I really like, and actually use on a daily basis, consist essentially of a simple list of headlines.

The simple presentation makes it easy to select the stories that interest me enough to read while still giving an overview of the rest.

While the 'follow a continuing story' idea is interesting (and the presentation here is impressive), it's not something that I value. I find that I follow stories naturally, and often across multiple sites and even media types. I don't really need to see it all in one place, all at once.


This is somewhat similar to a site I have been working on. http://www.newsdive.net I am doing filtering based on story content, but this seems to be based on tags and categories more.


Same question (as I posted in the main thread), if it possible to you to answer. Are you using machine learning, human editoring or both?


At the moment it's pretty simple matching based on a few rules, but machine learning is in development. In the case of google, it appears they are using human editors.


Thanks and good luck!



This is a pretty fun project. Nice, lightweight, combines multiple sources. Kind of embarrassing for the news outlets though that google can produce a better version of a full-feed experience than they can.



Interesting: "California software giant" is the tag the WP writer puts on Google.

(I put in a search to see what the net had on those words - seems mostly Oracle and a company with the SG name.)


I wonder if they are using machine learning, humans editors or both.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: