Tuesday, February 17, 2009

Graphical, contextual news

I'm sort of a news junky, in a loose sense. I was happy when all the major news sources began making their content available online, and I think it's convenient that secondary sources like news aggregators, RSS readers, and the blogosphere in general made accessing and organizing the massive datastream of online news significantly more manageable. Despite the advances in these distribution channels and the technology surrounding them, the way that news is presented seems archaic. It's as if the news medium is still boxed in by restrictions that were torn down long ago. I'm talking about the presentation of news information itself. I read Google News nearly compulsively because it's customizable, uses an effective (though fairly opaque) algorithm to choose stories of interest from credible sources, updates in real-time, and is extremely searchable. All these are excellent improvements on radio, television, printed news, and even online primary sources, but I think we can do much better--the basic form is still a better-organized portal to the same static news we've been reading since the beginning of news.

One of the frustrating aspects of online news is its overwhelming quantity. Despite the fact that I read at least the headlines and basic summaries of hundreds of articles a week, I find that I retain fairly few of the actual details, or even the bigger-picture concepts. I find myself editorializing as I skim the news stream: "another story about X" or "the Nth story this month about the debate on Y." Sure, I can choose the primary article on any given topic (as chosen by the Google computation) or I can choose one of a few thousand articles on the same topic from other sources. But, as I select a BBC story entitled Clerics urge new jihad over Gaza, I realize that what I'm really missing--and what even a background paragraph in this 558 word story can't adequately provide--is decent context. I don't want a link to the wikipedia article on the conflict in Gaza. I want to see how this article, and the 2,536 others like it, fits into the bigger picture. There's a ton of data surrounding any given news story, there are geographic locations, dates, names of political players, organizations, and pundits. Then, there are the hundreds and thousands of news stories leading up to this one; some of them are intricately related, and some are more subtly connected, but all help shed some light on what this particular story should mean to me.

I want all that context summarized for me and I think the best solution is a graphical one. I've seen news timelines, news maps, and other clever attempts, but none are focused on context. Those examples visualize and stratify news according to its meta-data.

If I were to design a graphical representation of news context, I would want to make sure that the high-level data was immediately accessible (with details available by drilling down). High school journalism classes teach that a good lead includes all of the vital statistics: who, what, where, when and why. This visual summary should provide the same, but with a scope that extends backwards in time, and outwards to related players and events that might provide incite into what this particular story means.

The Players. People, groups and organizations. Who is involved and why? Are they stakeholders or observers? What are the arguments being discussed, and have they changed over the history of this issue?

The Location. Is there regional context? What else is happening there right now?


The History. What were the events that lead up to this. Example: the Euro falls below $1.26. Show me where it was this time last year, upon its introduction in 1992, last fall when the US financial crisis hit, and its trend since. Where there are other spikes, show me the associations and correlations.

The Implications. What is likely to happen next? Have similar events occurred under similar circumstances? Example: flight 1549 to Charlotte, NC crashes after it strikes a flock of birds. Tell me about other recent plane crashes and the aftermath. Will US Airways market share shrink? Will there be an investment in bird-safe engines? Will air travel dip?

Having never consumed news in this way, I can't say for certain that such a reformatting of news data would help me retain the information that matters most, but linking small pieces into the "bigger picture" often helps to develop a more complete understanding.

I wanted to take a shot at this graphical approach, so I chose this article and did some related news search through Google to compile the big picture. I'm working on couple of formats which I will post here when I get a chance.

Combining this sort of graphical representation with technologies like Seadragon--introduced in this Ted talk (skip to 1:25 for the most relevant part)--could also be extremely effective.

Interacting with data in this way needn't be limited to the consumption of news either. It is, however, a good place to start and could be a great proof of concept. It seems to me that Google, a company dedicated to the accessibility of information, is extremely well-positioned to take another stab at it. With the ever-accelerating generation of new data, of which news is only a small slice, it seems to me that effective techniques at summarizing, contextualizing and interpreting this data will become more and more valuable.

No comments:

Post a Comment