The future of news. Right now.
Thank you Emi!!
This new Paul Ryan meme got us thinking about other political Tumblrs. We’re looking for good ones that we may not have heard of. So, if you know of any, please share. We asked readers the same question on WashingtonPost.com:
Tumblr isn’t a household name in politics. Yet.
For those of you who don’t know what Tumblr is, it’s — basically — a new(ish) blogging platform that allows seamless sharing of videos, links, quotes and almost anything else you can find on the Internet. (Check out our music Tumblr — it’s called Sir Fix-A-Lot — for an example of what it looks like. SELF PROMOTION ALERT!)
And Tumblr is starting to catch on in the political world.
Take the Tumblr “Hey Girl It’s Paul Ryan” , which came to our attention today and hinges on the the Wisconsin Republican’s efforts to eliminate the deficit and imagines that Ryan — like heartthrob actor Ryan Gosling — is able to make women swoon.
Generic shots of Ryan are paired with lines like “Hey Girl, let’s move forward with our relationship… by cutting spending and putting our love on a path to prosperity.” (That “Hey Girl” line builds off of a Ryan Gosling Tumblr that took off in 2010.)
The (Paul) Ryan Tumblr comes on the heels of “Texts From Hillary”, a Tumblr so popular that President Obama joked about it during his White House Correspondents Dinner speech on Saturday night .
The Fix — ever in search of the next big thing in politics — wants to hear from our community. Do you have a favorite political Tumblr? Do you run a favorite political Tumblr? Make your suggestions in the comments section and we will round up the best of the best in a separate blog post later this week.
1. What do you propose to do? [20 words]
Truth Teller — a mobile and desktop news application that captures, analyzes and fact checks events and speeches as they happen.
2. Is anyone doing something like this now and how is your project different? [30 words]
Crowd-sourcing and fact-checking are not new. But doing both in real time, through existing social networks, is unprecedented and would profoundly alter what leaders say and what people know.
3. Describe the network with which you intend to build or work. [50 words]
The application would convert political speech to text that could be parsed in real time to determine semantic intent and context that would be compared to the news organization’s data and Wikipedia with an algorithm to determine a level of truth. The truth level of the speaker’s sentence would be displayed on the live video as soon as the sentenced finished with a link to access the data that determined the truth level. Reaction from Twitter would also be displayed.
4. Why will it work? [100 words]
One of the biggest complaints about political coverage is that it allows untruths to go unchecked. This instantly dices the rhetoric and calls out statements that do not reflect reality. News organizations attempt to do a lot of this now manually. This would be an instant value in the newsgathering space and also add a real-time journalism layer to the increased world of in-the-moment news productions. For the mobile app build, it is convenient, immediate, simple and sharable and would reach corners of the world where other live streams may not be available to be checked.
5. Who is working on it? [100 words]
A cross functional team at the Washington Post that could expand, if needed, with funding.
Cory Haik, Executive producer, news innovation
Steven Ginsberg, National Political editor
Yuri Victor, UX Director
TJ Orentzi, Senior social media editor
6. What part of the project have you already built? [100 words]
The Post currently has a strong fact-checking brand on our site (http://wapo.st/zQed5i), and we already do crowd-sourced fact-checking around live events (http://wapo.st/xPHG3S), though these are manual and not in real time. We have seen success with engagement on these productions. The Post is committed to stretching this into a real-time, crowd-sourced and programmatic entity.
7. How would you sustain the project after the funding expires? [50 words]
The Washington Post would use Truth Teller for its own production and reporting purposes and would sustain the infrastructure of the product. But it is also envisioned as a tool to help journalists and others improve democracy across the country, and we would distribute it as widely and cheaply as possible.
Requested amount from Knight News Challenge:
Expected amount of time required to complete project:
Six months to a year
Total Project Cost:
Name: Cory Haik
Organization: The Washington Post
We have an idea and we need your help to make it a reality. We want Knight to help fund a project that would take politicians’ rhetoric and fact-check it in real time.
To win funding from Knight, we need to prove that there’s an interest in the project. Help us prove that the truth matters (and that this project has potential) by reblogging this — and sending along your suggestions to make this project better.
By Emi Kolawole
If you’ve been following the Post’s political coverage, you’ve probably come across @MentionMachine. The tool, which tracks mentions of the 2012 presidential candidates on Twitter and in the media generally, launched Tuesday and is the brainchild of Executive Producer Cory Haik.
I spoke with Haik Friday morning to discuss how @MentionMachine came to be and the impact it might have on election reporting.
"Well, we’d been talking about Twitter and the elections since I’d walked in the door of the Post,” said Haik, who started a little over a year ago as deputy editor of the Post’s Universal News Desk. She has since become Executive Producer for News Innovation and Strategic Projects.
Concepts can either fizzle or thrive depending on available resources. In this case, Haik had already chosen a name. “I feel like once you give something a name, it becomes a thing,” she said. “I squatted on the Twitter account long before it was even an official project, because I thought, 'This thing's gonna' work.'”
After a series of memos, meetings, and presentations on a wash-rinse-and-repeat cycle, Haik started to bring newsroom resources to bear on the project, teaming up with Washington Post National Digital Editor Amanda Zamora, Developers Jesse Foltz and Sean McBride, Designer Katie Parker and PostPolitics Social Media Producer Natalie Jennings.
What resulted is a tool that is more complex than the otherwise simple menu bar would indicate. Click on a candidate and you can see how often they are mentioned on Twitter relative to other Republican primary candidates and President Obama.
But wait, there’s more.
"We keep talking about Twitter, but we’re also counting Trove," said Haik of the Post-acquired company that brings in more than 10,000 publications, targeting them to users’ preferences. “So, that’s an awesome win there.”
Jennings reports on the @MentionMachine data regularly, mining it to help inform the Post’s journalism. ”The daily data is proving itself,” said Haik, “and doing that reporting is a big part of the machine. It’s not just the numbers.”
The @MentionMachine has a profanity filter, but tweets with profanity are counted among the candidates’ mentions. Try to get any more specific than that, and you run into what Haik calls “a special sauce” — the programming muscle that powers @MentionMachine. Haik says she and others refer to it as a combination of Twitter, her own unique innovations, and the execution engine Kangaroo.
Registered voters are not given preference, since @MentionMachine does not target those in the U.S. or provide U.S. geotagged tweets any additional weight. @MentionMachine does, however, feature “Top Tweets” — or tweets that are considered to be driving the conversation about a particular candidate at the given moment. “We have a very unique formula algorithm for top tweets, it’s not just retweets,” said Haik, emphasizing that there are safeguards in place against individuals trying to game the system.
"We’re making sure that our top tweets are tweets that are driving the conversation,” she continued, “So, the point is we’re trying to show people exactly where people are spiking and when.”
The Post is not the only organization trying to mine what has come to be known as “big data” when it comes to the 2012 election. CNN debuted a similar Twitter tracking tool on Iowa caucus night, and other Web sites are drawing from the Twitter API data pipeline as well. Regardless of what others are doing, Haik said she’s confident that @MentionMachine has carved a unique place for itself as a transparent tool able to provide journalistic value. “We tried to make the machine programatically do some journalism for us,” she said.
While @MentionMachine shows candidate mentions, Haik is often asked about sentiment data — or whether a tweet or media mention can be classified as falling into one or more emotional categories, such as sarcasm, anger, sadness, and joy. “Sentiment data’s not there yet,” said Haik, “I don’t feel like you have to have sentiment data to get a good read on where the
conversation is on a person.”
The @MentionMachine is by no means static. It is continually tweaked and improved. “I’m calling it a very mature prototype,” said Haik. She provided a preview of one impending change: “What you’re going to see next week is when you hover over that spike tweet, you’re going to get that exact top tweet for that moment.”
Read more about @MentionMachine and explore the data.
Facebook recently released a list of its top 40 most shared news stories this year, which includes three Washington Post stories. But why should Facebook have all the fun?
Here are the most popular Washington Post stories this year among Tumblr users, based on the number of clicks from the Tumblr site. Take a walk down memory lane and remember the year that was.
This week we launched the Better/Worse Life Project, which looks at population growth, unemployment and median income across the country over the past 30 years and uses that data as a background to frame the very simple, open question ‘Is life in your state getting better or worse?’
It uses your IP address to automatically detect your county and state so that we can ask you about the area that’s most relevant to you at the moment.
After we have your vote, we ask for a couple of simple demographics – age and race – and display a visualization that shows where your state falls on the scale from better to worse based on the votes we’ve gathered. You can then filter the responses based on race and age, or sort the states based on unemployment, median income or population growth, to see if there is trending based on these metrics.
This is one of the first projects to take users opinions and mash them up with actual data to see whether perceptions match up with reality. It’s a fascinating window into how people feel about the places where they live and a forum for a conversation around how things are changing.
There are already some patterns emerging. We are starting to see that, in general, more states with high unemployment are being rated ‘worse’ by users, and more places with low unemployment are being rated ‘better’.
D.C. stands out – though unemployment is high, 83 percent of users (as of publication of this post) ranked it better. One left a comment after rating it ‘better’: “DC is better due to a higher number of permanent residents, community activism, and better stewardship. Welcome to the 21st century!” The comments on why people voted the way they did have been some of the most interesting results of the project we’ve seen so far.
This kind of presentation can be risky. It’s so dependent on user feedback that if no one participates, there will be nothing interesting to look at. But it’s worth the chance – the kind of fascinating information we can gather once it does get going is possible only when you open the doors to participation from users. We’re collecting information on counties in addition to states so that, if we get a lot of responses, we can display a profile of the state that shows whether residents think counties are getting better or worse. It is going to take a lot of responses to get feedback for over 3,000 counties, but it might be possible with your help.
Right now, we have about 4,000 responses, but we need many more. Rate your state here: http://wapo.st/betterworselife
- Kat Downs / Innovations Editor for Graphics
Q: Why is the WaPo so consumed with this case? Why are you not treating every lost life with such importance as you are with this one? Yes, a life was lost and that is tragic within itself, but what’s even more tragic is how the WaPo and the rest of the media plays up certain cases because of the zip code and skin color of the people involved. Many people are killed in this region every year, where is their coverage? — kdofour2000
Local blogger Katie Rogers has the answer on Ask the Post.
The Washington Post’s new Primary Tracker, created by Kat Downs, Ted Mellnik and Karen Yourish of The Post’s Graphics Department, is a one-stop shop for keeping up with the race for the Republican presidential nomination.
The interactive graphic features the newly-finalized primary schedule, state visits by candidate, and historical data from previous GOP presidential primaries. The tracker pulls state visits from PBS NewsHour’s political calendar and aggregates the data by state and candidate. Live results for each primary will be available as they happen.
"One of my jobs as a multiplatform editor at The Post is to put together the corrections for the print edition. It’s a bit of a grim task, staring failure (sometimes my own) in the face, but occasionally I see a correction and can’t help but smile. For instance, one correction in today’s paper was straight out of a ‘Seinfeld’ episode.”
Continue reading at Ask the Post.