Two months ago I made a post about the fun little news application on the Nintendo Wii. Dan Burd responded to the post with this comment criticizing some of Wii News’ interface assumptions: “I think it’s limiting to say that each news story only pertains to one location. Many news stories are overviews of the relations between two or more countries. I’m guessing the AP thing would place them at whatever city the reporter is reporting from. I think that’s a bit misleading.” If you ask me, he is spot on.
Burd’s comment refers to global news, but the meaning also applies to other scopes. To use his terminology, it is limiting and potentially misleading for any news interface to require that each story pertain to exactly one location.
To prove the point, here are some abstracted stories that wouldn’t fit well into such a restrictive schema:
- Events that take place in one area but are directly relevant to another.
- Interactions between multiple communities.
- Any situation where affected location and occurring location are different. (This includes stories with global relevance).
- Related events that happened in more than one place.
The list could go on, but you probably get the picture.
In database speak you would call the desired relationship between news articles and locations as “many to many,” meaning one article can have many locations, and one location can have many articles. From a programming standpoint, it is simple enough to address that relationship when designing your software; it’s just a matter of recognizing that it’s there.
Since the technical implications of having multiple location tags aren’t too severe, I bet a lot of people already do it (it isn’t an advertised feature on most geocoded news sites as far as I can tell, so I don’t know for certain). For anyone that doesn’t, though, I’m bringing it up to put it on the radar as a really important feature for any news site that uses location information to categorize and display its articles.
A quick warning: adopting this slightly added complexity means you have to be sure not to confuse the reader. (If you are using the map as a way to help users filter news instead of browse, then you are probably pretty safe.)
(This post pertains to a bullet point from Tapping the Potential of Geotagging – Include news that is tagged to multiple locations, irrelevant locations, or no location at all.)
I am always spot on
You hit the one point I was going to bring up– that the data model isn’t the problem, it’s the user interface.
But what about an overview of a town where an action might apply to a protest at town hall about a development a mile away?
Even used to filter news, how do you prevent this from becoming noise? Does the story have the same symbol, same pop-up summary, same link in multiple locations?
Yeah Ben,
I admit I kind of sidestepped the issue :); one reason for that is that I’m going to dedicate a post to interfaces. The other reason is that good interfaces are tougher to think up (for me anyways!) and interface is a much more “case by case” kind of thing – each site will have its own interface requirements and tradeoffs so there is no overall “perfect interface”.
I can say right off the bat that I think it will take more than just displaying two points on a map (Like you imply that approach will tend to indicate to the reader that there are two separate stories.) Even if the descriptions are the same or they have the same symbol there are probably better ways to do it.
I guess I’ll just ask for a rain check until that interface post.
There is very active discussion going on right now on the GeoRSS mailing list about how to handle just this concept of multiple locations: from a data interchange mechanism, use cases, and also interfaces.
There is some information here GeoRSS Multiple Locations, that would be useful to have feedback on.