Articles

If you present everyone who reads the a newspaper online with a custom, personally generated version of the paper you will end up giving people a nagging "what am I missing out on" feeling. This feeling will lead people astray as they fight to make sure that they are accurately being informed. They will not trust the machine to get it right.

People are not always in the same frame of mind (see Yahoo! Mindset search) and they're not always interested in the same topics. Personalization technology needs to address this variability if it's ever going to be adopted widely.

I wrote about this in greater detail a year ago:
http://breasy.com/blog/2007/04/23/the-limits-of-personalization-technology/

I guess, totally personalized news will co-exist vis-a-vis generic newspapers. Maybe different pricing models will come into the fray, cheaper ones are generic, Warren Buffet gets his version and you pay $1000 per month to read the Warren Buffet's version of NYT.

Moreover, all newspapers are not about news some are pure propaganda play. Those could really become saavy and bend the rules/algorithms to their advantage. Imagine the leader of the right-wing party subscribes to a newspaper that is run by the left-wing or say the mafia. He would rather pick his news from a platter than getting something he never believes.

Impeccable Technology + Correct Pricing Models would drive the show going forward, IMHO.

I do not think personalized news works for all news requirements.

Eventually, a wise reader knows that to think outside the box, s/he has to get outside the box, and that means reviewing interests they know nothing about.

For those with matured interests and need for constant updates, a personalized feeder system saves from searching news for known interested issues, such as Dow Jones, advances in ONE disease, or a missing person report.

A key issue about news is not personalized news, but the quality and bias free honesty of news.

IEEE Member

Echoing the comment by Rajiv Das, there was an MIT research project named "Fishwrap" in the 90's focused on developing "personalized news". One of the interesting lessons learned addressed the social and cultural implications of personalized news: [1] serendipity: in a recommendation-based world, is there value in inducing people to read about topics they would otherwise have missed or had not previously expressed (related) preferences for?
[2] Balkanization: is there a danger of further self-reinforcement if people read only the news and commentary for which their sympathies are predisposed.

IEEE member