Monday, May 15, 2006
PERSONALIZATION: WSJ piece on news personalization services
ORIGINAL URL:
http://online.wsj.com/article/SB114727164857848939.html
PUBLISHED: May 15, 2006
THE JOURNAL REPORT: TECHNOLOGY
Me, Me, Me
The personalized newspaper was dreamed up two decades ago. We're getting
closer and closer.
By JESSICA MINTZ
May 15, 2006; Page R9
The Web has made accessing far-flung news outlets, from tiny local papers
to major foreign presses and every Web log and magazine in between, as
simple as a couple of mouse clicks. But keeping that deluge of information
organized hasn't been so easy for readers.
Now, some Web sites are taking a stab at a new solution:
more-sophisticated personalized news pages. These sites are the latest
step in the evolution of "The Daily Me" -- the name coined by the founders
of the Massachusetts Institute of Technology's Media Lab two decades ago
for their idea of a perfectly personalized newspaper.
But those tools have a couple of major shortcomings. One is that they do
nothing to help readers deal with the sheer volume of material that's out
there: It's not uncommon for people to track dozens of blogs, news sites
and other Web sites with RSS feeds, and much of the material they collect
this way inevitably goes unread. Then there's the other side of the coin:
Readers may be missing out on lots of other news they'd be interested in
but that doesn't make it through their filters.
Now, a new generation of Web start-ups is trying to address those
drawbacks in different ways but with a common premise. These sites track
the reading habits of their users as a whole, then use that data to make
suggestions to individuals based on what others like them are reading.
This communal news judgment can help ensure that readers don't miss
important stories outside their usual interests. And it can even help
online news junkies decide which of the stories they choose to see are
must-reads, and which can more safely be skipped.
One of these new sites is Rojo.com3, owned by San Francisco-based Rojo
Networks Inc. Rojo.com helps users find and organize RSS feeds. Instead of
just having users plug in the feeds they know they want to read, Rojo asks
new users to define their general interests and favorite sources first.
During the sign-up process, users check boxes next to topics such as "News
-- top stories" and "Iraq/military bloggers." Then they can choose from a
list of big-name news sources. Rojo automatically subscribes new users to
popular feeds that match their interests. Readers who already had a list
of RSS feeds can have those displayed on the site as well.
Users can add a new feed to their Rojo account at any time or use the
search box on the site to find new ones; a search for "iPod," for example,
brought up 67 different feeds.
What really distinguishes Rojo from other RSS readers -- sites that
organize and display RSS feeds -- is the way it uses data it collects from
all of its users, currently about 100,000 unique visitors per month. The
site looks at each individual user's feeds and interests, and figures out
what's missing from his or her list that similar members are reading. In a
box at the top of each page, the site recommends one new feed every time
the user clicks onto a new page.
Rojo members can view their feeds by "relevance," in addition to date or
alphabetical order. The relevance ranking automatically combines the
user's personal interests and past reading patterns with how "hot" the
story is among other readers -- especially readers with similar profiles.
So even if a reader doesn't take advantage of the site's suggestions for
new feeds, it can help that reader prioritize the feeds he or she is
receiving.
Michael Davidson, co-founder and chief executive of Seattle-based Newsvine
Inc., says he follows more than 100 handpicked feeds, mostly about the
technology industry, in an RSS reader. Still, he feels like he could miss
out on important stories. "How am I supposed to find out about bombing in
Iraq, or hunger in Africa," he asks, when "I haven't specifically said
'Yes, subscribe me to those topics'?"
Newsvine.com10 tries to solve that problem by relying heavily on the
collective behavior of its users. The site, which launched to the general
public on March 1, is built around content from the Associated Press wire
service and Walt Disney Co.'s ESPN11. But the layout isn't static. The
actions of all Newsvine users push certain stories to the top of the
site's main page, and to the top of its sub-pages (U.S. and world news,
technology, sports, etc.). In addition to just reading a story, Newsvine
users can click on an arrow icon to vote on its importance. They can also
make comments and join live chats about each story. All of those factors
together affect how prominently a story is displayed on the site.
The typical Newsvine page leads with a top news story, displayed as a
headline, splashy photo, first paragraph and link to the complete story.
It is followed by a short list of the most voted-on and most commented-on
stories; the latest top articles from the AP wire (ranked by AP editors)
are shown as smaller links off to the side.
Parts of the page are set aside for "seeds," or stories from blogs or news
sites outside of Newsvine that users recommend. Those articles rise to the
top of the list the same way news stories do -- through votes and comments
of readers.
Users can also personalize their pages by setting up "watchlists." Every
news article, "seed" and member home page (which doubles as a blog) gets
tagged with keywords. Users can search for a keyword and, if it exists,
click on a green button to "watch" it -- the equivalent of subscribing to
an RSS feed of every story with that tag.
For example, a member could have searched for "Syracuse" and then added it
to a watchlist to follow the surprising victory of the university's men's
basketball team in the Big East conference tournament, and the team's
quick exit from the national tournament.
A third news site, Findory.com12, also relies heavily on the behavior of
its users, but doesn't require them to list their interests, select feeds
or vote on stories. Instead, it works on the same principle as
Amazon.com's recommendations, which is no coincidence: Findory.com Inc.
was founded in 2004 by Greg Linden, the engineer behind Amazon's
recommendations engine.
Findory, which has amassed a following of about 100,000 visitors a month,
looks at an individual's reading history, compares it with similar
readers' tastes, and offers up links to stories that similar readers have
enjoyed -- or at least read.
The first time a user visits Findory.com, the home page shows a mix of
news stories and blog entries, based on a combination of general
popularity and how recently the stories first appeared. Findory assumes
that when a user clicks on a story, he or she is interested in the
subject. Each time the user returns to the Findory home page after
clicking on an article, he or she will find the page reconfigured with a
different mix of stories, some marked with sunbursts to indicate a
recommended read.
--Ms. Mintz is a staff reporter in The Wall Street Journal's New York
bureau.
Write to Jessica Mintz at jessica.mintz@wsj.com13
A sampling of other innovative news Web sites:
. Reddit.com
Links to articles on the Web, submitted and ranked by votes from the
site's users. Every time you vote for a story, the site gets to know what
you like and filters results to suit your tastes.
. Digg.com
Technology news links, submitted and ranked by reader votes.
. Memeorandum.com
Links to the political stories generating the most buzz on the Web; draws
together mainstream news articles, blogosphere discussion and related
links.
. Feedster.com
A directory of sites that send out updates to subscribers using a
technology known as an RSS feed. Type a keyword into Feedster's search box
and find feeds from blogs and news sites.
. PubSub.com
The site trawls millions of data sources for keywords that match your
interests. When a new match is added, it tells you.
. Backfence.com
An experiment in hyperlocal news written by residents of a community.
Topics range from weekend events and local business ratings to municipal
news stories. Available in just a few communities so far.
----------------------------------------------------------------
The article above is copyrighted material, the use of which may not have specifically authorized by the copyright owner. The material is made available in an effort to advance understanding of political, economic, democracy, First Amendment, technology, journalism, community and justice issues, etc. We believe this constitutes a 'fair use' as provided by Section 107 of U.S. Copyright Law. In accordance with Title 17 U.S.C. Chapter 1, Section 107, the material above is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. If you wish to use copyrighted material from this blog for purposes beyond fair use, you must obtain permission from the copyright owner.
