Tumblelog by Soup.io
Newer posts are loading.
You are at the newest post.
Click here to check if anything new just came in.

February 23 2011

17:00

The context-based news cycle: editor John O’Neil on the future of The New York Times’ Topics Pages

“There’s are a lot of people in the news industry who are very skeptical of anything that isn’t news,” says The New York Times’ John O’Neil. As the editor of the Times’ Topic Pages, which he calls a “current events encyclopedia,” O’Neil oversees 25,000 topic pages, half of which — about 12,000 or so — include some human curation.

While the rest of the newsroom is caught up in the 24-hour news cycle, constantly churning out articles, O’Neil and his team are on a parallel cycle, “harvesting the reference material every day out of what the news cycle produces.” This means updating existing topic pages, and creating new ones, based on each morning’s news. (The most pressing criterion for what gets updated first, O’Neil said, is whether “we would feel stupid not having it there.”) A few of the Times’ most highly curated topics include coffee (curated by coffee reporter Oliver Strand with additional updates by Mike White) and and Wikipedia (curated by media reporter Noam Cohen),  as well as more predictably prominent topics like Wikileaks and Egypt.

The Topics team includes three editors and two news assistants, who work with Times reporters. “People give us links to studies they’ve used for stories or articles they’ve looked at, and this is something that we do hope to expand,” O’Neil said.

But half of the topic pages are “purely automated,” O’Neil said. And O’Neil is even contemplating contracting the number of curated topic pages, as people and events drop out of relevance. (The Topic pages garner 2.5 percent of the Times’ total pageviews.) O’Neil said he had read a statistic that roughly a third of Wikipedia’s traffic came from only about 3,000 of its now more than 17 million pages. “We’re concentrating more on that upper end of the spectrum.”

In a phone conversation, I talked with O’Neil about why the Times has ventured into Wikipedia territory, how the Times’ model might be scalable for local news organizations, and why creating a “current events encyclopedia” turns out to be easier than you might think. A condensed and edited version of that conversation is below.

LB: How did the topic pages develop?

JO: Topic pages began as part of the redesign in 2006. Folks up in tech and the website realized they could combine the indexing that has actually gone on for decades with the ability to roll up a set of electronic tags. The first topic pages were just a list of stories by tag, taking advantage of the fact that we had human beings looking at stories every day and deciding if they were about this, or were they about that. Just that simple layer of human curation created lists that are much more useful than a keyword search, and they proved to be pretty popular — more popular than expected at the time.

LB: What’s the philosophy at the Times behind the topic page idea?

JO: Jill Abramson’s point of view when she started looking at this: When she was a reporter, she would work on a story for days on end, weeks on end, and pile up more and more material. You end up with a stack of manila folders full of material, and she would take all of that and boil it down to a 1,200-word story. It was a lot of knowledge that was gained in the process, and it didn’t make it to the reader. The question was: How can we try to share some of that with the reader?

My impression is that people find these pages terrifically useful. Not everybody comes to a news story with the knowledge you would have if you’d been following the story closely all along. News organizations are set up to deal with the expectation [that people] have read the paper yesterday and the day before.

LB: How do you go about transforming news stories into reference material? What does the process look like?

JO: What we found, as we did this, is that the Times is actually publishing reference material every day. It’s buried within stories. In a given day, with 200 articles in the paper, about 10 percent reperesent extremely significant developments in the field. Now we can take a small number of subjects, like Tunisia or Egypt or Lebanon or the Arizona shootings, and keep on top of everything, set the bar higher. We can really keep up with what the daily paper’s doing on the biggest stories.

LB: As you note, there’s a lot of wariness among “news” folks  around putting  effort into topic pages. For instance, when I talked with Jonathan Weber of The Bay Citizen, the Times’ San Francisco local news partner, about topic pages, he told me: ”people are looking for news from a news site….We’re not Wikipedia. You don’t really go [to a topic page] for a backgrounder, you go there for a story.” How would you respond to that?

JO: Our experience has been that that’s never been entirely true, and it’s becoming less true all the time. Look at the pound-and-half print New York Times, and think how much of that is about things other than what happened yesterday. Even in the print era, that was a pretty big chunk.

Then again, it makes sense for folks at a place like The Bay Citizen to be more skeptical about topic pages. A blog, after all, is all about keeping the items coming. And a site focused on local news would feel less need to explain background — hey, all our readers live here and know all that! — than if they were covering the Muslim Brotherhood, for instance.

LB: So what about the Wikipedia factor? Why should the Times be getting into the online encyclopedia business?

JO: I think Wikipedia is an amazing phenomenon. I use it. But there’s no field of information in which people would find there to be only one source. On Wikipedia, there’s the uncertainty principle: It’s all pretty good, but you’re never sure with any specific thing you’re looking at, how specific you can be about it. You have to be an expert to know which are the good pages and which are the not-so-good ones.

Our topic pages — and other newspaper-based pages — bring, for one thing, a level of authority or certainty. We’re officially re-purposing copy that has been edited and re-edited to the standards of The New York Times. It’s not always perfect, but people know what they can expect.

LB: What’s the business-side justification for the Topic Pages?

JO: We know that the people who come to the topic page are more likely than people who come to an article page to continue on and look at other parts of the site. It helps bring people to the site from search engines.

It’s also brand-building; it’s another way people can form an attachment. People can also subscribe to topic pages. (Every page produces an RSS feed.) We’ve begun to do some experimenting with social media. There are lots of people who want to like or follow or friend The New York Times, but a topic pages feed gives you a way of looking at a slice of this audience. It turns the supermarket into a series of niche food stands, so to speak.

LB: The Times obviously has a lot more resources than most local news outlets. Is developing topic pages something of a luxury, or is it something that makes sense to pursue on a more grass-roots level?

JO: At the Times, less than one half of one percent of the newsroom staff is re-purposing the copy. That makes it of lasting value, and makes it more accessible to people who are searching. If you think about a small regional paper, three editors would be a huge commitment. On the other hand, the array of topics on which they produce a significant amount of information that other people don’t is small. There’s a relatively small number of subjects where they feel like, “We really own this, this is key to our readership and important to our coverage.”

If people think of topic pages as the creation of original content on original subjects, it never looks feasible. If you think about it as re-purposing copy on your key subjects, I think it’s something more and more people will do.

September 14 2010

20:00

More growth for Gawker comments, and more power to elite commenters

We’ve written before about the commenting system at Gawker Media’s family of sites, which for my money strikes the best balance between complexity and simplicity, between encouraging good behavior and policing bad. (I also just like that Gawker’s a company that really thinks about comments, that doesn’t just treat them as an expected annoyance/pageview driver.)

The chart above shows the growth Gawker comments over time; I’ve highlighted the section on the right that represents the continued increase since I last posted these numbers in April.

So it’s noteworthy that Gawker announced today a couple small tweaks to their commenting system.

First, they’ve added a few more gradations to the kinds of discipline available to wayward commenters. Before, commenters could be banned, and individual comments could be disemvoweled (rendered less legible by removing the vowels — although some would argue disemvoweling does more to draw attention to the bad behavior than it does to punish it). Now, commenters can also be officially warned for straying from proper behavior (with a link to official commenting policy) or suspended for a week. I’d imagine that these lesser punishments might discourage bad commenters from going through the bother of creating a new false identity and continuing to stink up the joint. And it could also help people who genuinely don’t realize they’re being bad.

Of note is that these power don’t just rest in the hands of Gawker Media staff: These tools are also available to the army of starred commenters who have impressed Gawkerites with their work. So here, for instance, Gawker user morninggloria has warned a commenter for daring to say Lady Gaga looked like John Lennon in drag. (I’d like to thank morninggloria for giving me an excuse to create a “John Lennon in drag” tag here at the Lab.)

That kind of decentralization makes it tenable to govern the huge crush of comments these sites get, and it also sets a goal that encourages good behavior: write enough good comments and you’ll get a gold star and some authority to shape the site you love.

The second major change is what they’re calling thread moving. Here’s an explanation from Gizmodo’s Jason Chen:

Then, there’s thread moving. That’s what we do if we think a comment is so egregious that it deserves both a warning and a moving to a tagpage, so it’s not cluttering up the discussion. Here are the main five tagpages we’ll be moving to.

• #trollpatrol. Originally we used this tag for identifying trolls, but we’ll throw actual trolls in there as well. But please, continue using that as a place to show us where the trolls are.

• #fanboys. Another obvious tagpage. This isn’t for people who use and enjoy products, it’s for people who lose their damn minds over a brand or idea and are blind to any other options or dissenting opinions. You should know who these are.

• #timeout. A place where we send commenters that need a little time away from typing words into boxes in order to think about whether or not this is the right place for them. This goes with a 7-day suspension—something milder than a ban, but still serves the purpose of telling them that we don’t like what they’re doing with their comments.

• #phantomzone. If you make uninformed, stupid or otherwise lousy comments, this is where that comment will be. Say hi to Zod.

• #whitenoise. Offtopic discussions go here. If a post is about keyboards and you talk about picking out new curtains, we’ll escort you over here.

• #dev/null. I just came up with this one, and I’m pretty sure I’m going to use it quite often.

Essentially, it’s a way to apply tags to individual comments, have them detach from their native post, and live another life in the Gawker Media forums, which are tied to the tags. This could separate off-comment topics without killing them off completely; one suspects the bad-behavior tag pages will have their own regular denizens. And the change could work to liven up the tag-based forum pages, which it appears have gained traction in only a limited number of cases. (See Deadspin’s #iwasthere tag page, or Gawker’s #tips page.) I love the concept of comment tag pages — treating the comment as an independent unit of content, opening up new avenues for involved commenters to create and contribute — but I’m not sure how well it’s worked in practice. It may be the point where the system grows too complex for most users.

Gawker Media CTO Tom Plunkett posted the above chart today and added this about the changes:

At Gawker Media, comment growth continues to be strong — both in volume and quality. It’s good to see validation of the processes we’ve introduced.

This week we’ve rolled out new features that will allow us to further improve the experience. It is now possible to move comment threads from posts to forums (think “off topic” threads: we’re happy to let you keep the conversation going, but it’d be better to continue the discussion in a forum appropriate to the subject). We will utilize thread moving for many situations (off topic, inappropriate comments, bannable offenses, etc.), and think it will only improve our platform.

If you are paying attention to beta.gawker.com or beta.jalopnik.com, you will see more improvements we plan to roll out in the future. Remember – these sites are beta (alpha may be a more appropriate description)! Don’t expect everything to work perfectly all of the time!

So if you want to see what Gawker’s thinking about for the future, the beta site features a more magazine-like front page (as opposed to straight blog hierarchy — the most popular recent story gets top billing), non-standard fonts via Typekit, a wider story well, smooth page transitions, a stationary sidebar, and a more prominent footer. We’ll see how much of that reaches the production sites of one of the more adventurous new media companies around.

April 05 2010

16:03

Build Your Own NYT Linked Data Application

Learn how to build an application with linked data from The Times.

November 06 2009

00:18

Google News embraces self-identification of content

Some online-only news organizations were upset when Google News began attaching a “(blog)” label to their content two months ago. Others, like me, complained the label was outdated and inconsistently applied.

Now Google News is asking publishers to label themselves. In an update to its sitemap standards announced today, Google News is requesting that sites explicitly tag content that’s published on a blog. Same goes for press releases, satire, opinion, user-generated content, and any articles that require registration or payment to read. The technical details are here.

Most of those labels will be visible to users of Google News, as they are now. Opinion and user-generated content won’t get a label but will presumably affect search results. And while tagging is voluntary, Google reserves the right to “add such designations to certain articles as necessary.”

I still don’t see why it matters if news is published on a blog or some other platform. (Google CEO Eric Schmidt ventured a distinction yesterday.) But allowing publishers to self-identify their content is a big improvement that should resolve most of the complaints Google News has been hearing — and which have been voiced to me in private. It’s a small issue with much bigger implications for how we consume, sort, and, yes, identify news in the future.

Older posts are this way If this message doesn't go away, click anywhere on the page to continue loading posts.
Could not load more posts
Maybe Soup is currently being updated? I'll try again automatically in a few seconds...
Just a second, loading more posts...
You've reached the end.

Don't be the product, buy the product!

Schweinderl