Moby Dick Project

Moby Dick Project

UPDATED INFO: Our overall mission is shaping up to be: helping "power-bloggers" -- by which we mean people who blog professionally, frequently, and work within some sort of community (either via direct employment or freelance relationships). Specific things we want to improve are the fact-checking and editorial processes, and commenting (consumption & production thereof). Our target users have existing behaviors that we want to augment (not replace) with tools that will increase a) the quality of their output and b) engagement with their audience & colleagues. Some insights we've had along the way are:

  * News-writers working exclusively online (aka bloggers, as opposed to traditional journalists) are more likely to understand and be receptive of the potential benefits provided by improved tools & platforms 
  * No one is going to adopt technology that demands they completely change their existing habits
  * Common practices like exchanging drafts and edits with email and .doc attachments (or preview-links) are pervasive, so providing something distinctly better that doesn't interfere with existing, linear workflows is key.
  * Evangelism is key for adoption, which means prioritizing ease of installation, ease of use, and shallow learning curve -- all of which point to minimal, incremental improvements

With those things in mind we're designing around a set of constraints:

  * Any new tool must integrate with well known, popular platforms, and not look completely different than what's already out there -- we have Wordpress plugins in mind, specifically
  * Emphasize collaboration (even just the potential for collaboration) -- we think of GitHub as a good model for this. A "GitHub for news" would necessarily look and work very differently than GitHub itself, due to fundamental differences between coding & news/storytelling, but the ideas of being able to "diff" facts between articles and collaboratively build stories in a distributed manner ("branch", "merge", "blame") are most exciting to us.
  * Always reveal the times at which things happen (whether reported events or updates to facts within an article)
  * Make use of annotations & edits within the text of an article, keeping in mind existing metaphors of hyperlinks and margin-notes

Sticking points so far include:

  * Whether to initially approach a solution from an interface or platform standpoint. One one hand there are lots of great interface ideas out there, but they need a platform. On the other hand a new platform for transparency and accuracy in the news could be revolutionary, but not if there's no interface for people to use it. (Classic chicken-and-the-egg)
  * Whether to treat this as a standalone project with core development team building things for a group of users, or to solicit involvement from an ecosystem of sympathetic contributors around a coherent set of specifications and principles. (And does such an ecosystem exist already that we could leverage?)

The Idea

We've gone down a few roads while prototyping, ranging from what a “Salesforce for journalists” might look like to the idea of a semantic ontology for news. Our core idea is that bloggers and journalists on deadline need more expedient access to a high-quality global knowledge base.

...which we're sticking with, but the precise users we're going after now are professional bloggers (perhaps freelance) who write several columns/posts per day, use inline hyperlinks excessively, and are read widely enough to garner a significant amount of comment/Twitter/Facebook chatter.

We've been expanding on the Back-Story and Data-Check + API ideas from the summer d.school session (and are considering how reader observations might be made to work harder for journalists). We think that over-worked journalists on deadline need more expedient access to a high-quality global knowledge base (an idea that we're discussing with journalists), and we're exploring the following:

How can comments (which are largely forgotten about at the end of current stories) better point to what's working with a story and where more information is needed? Would a journalist-only tool for marking comments and tweets for future exploration be helpful to journalists? Could this create a pipeline for related stories? How do trusted commenters' contributions bubble up? What role does sentiment analysis play?