I agree with your assessment that datelines aren’t geocoding. But as it stands currently, it’s the next best thing that newspapers have in adding location information to stories online (if, as you said, they even put datelines in stories).

Ellington, the CMS used by Scripps’ newspapers, allows for custom datelines if producers take the time to add them to stories. Most sites simply let the feed from print carry them directly into the story.

Our CMS also has a places application that could theoretically allow for a very detailed geocoding (lat/long, etc) of stories. But again, this would be a change in workflow for newsrooms.

And that, I believe, is going to be the largest hurdle to leap in getting stories categorized in this manner.

Also, I suspect there would be some confusion as to how to tag certain stories. What location does a profile story belong to? What about a feature that spans multiple locations? Or a story that has no location, such as a column?

Patrick, I should’ve been a bit more specific. What Topix does –as I understand it– is scan for keywords that gives a sense of what locations pertain to the story. So if the dateline has “Maitland,” Topix is going to send that to the city of Maitland news page.

On the contrary, geocoding a story means drilling down to the exact latitude and longitude, which can only be done by a human or by processing an address (or zip code, more generically). Plotting those geocoded stories on a map means a user could potentially type in his address and get all the news stories that have taken place within, say, a 20-block radius. In big cities, tagging a story “Miami” or “Los Angeles” is almost useless for local news purposes.

Many, many newspaper sites have been been metatagging locations for years.

It’s quite a leap, however, from location tagging to true geo-tagging, with specific lat and long. That requires processes and/or software of a magnitude more sophistication.

Simply adding a field that says “New York” is not geo-tagging.

Plus, datelines are a funny thing. None of the newspapers I’ve worked at dateline stories from their home city — their most valuable and unique content.

Most of our content — both news stories and user-contributed content areas such as blogs and user submitted stories — is available in RSS feeds, which is increasingly the case at many news operations.

We have a form in our content creation tools that lets users enter an optional address, and that causes a Google map to appear in their content. We found out that we could put the geocode for that address into the RSS feeds. Heres one example from a blog on Bakersfield.com:

The blog with maps: http://people.bakersfield.com/home/Blog/pachecod
The RSS feed: http://people.bakersfield.com/home/Blog~/pachecod

Here’s how that map at the top is geocoded in the RSS feed:

35.374448
-119.019958

We had no idea what adding the geocodes to RSS feeds would do, but then a few months later one of our more adventurous staffers discovered that Google Maps will create create a map of posts for any RSS feed that contains geocodes! You just paste the RSS feed into the Google Maps search box. What you get back is something like this:

http://maps.google.com/maps?f=q&hl=en&geocode=&q=http:%2F%2Fpeople.bakersfield.com%2Fhome%2FBlog~%2Fpachecod&ie=UTF8&ll=26.902477,-92.373047&spn=40.183942,58.535156&z=4&om=0

(Here’s a hyperlink if that URL got cut off in your mail reader)

Let’s hear it for standards! I’m curious to know of other tools that can process geocodes from RSS feeds in this way.

-Dan

It’s not very granular, but if you adjust your CMS to accept the dateline as a field in your database you have effectively geocoded your story. Of course, you’ll also need to create something like Ellington’s generic places app so all content types can be given a location as well.

The biggest problem here is that the newsroom is still entering data into a print system that may not allow for the export feed to be parsed in this manner. In this case, it might be a little extra work for folks to add the location data.

This means there should be a revenue potential for use of geocoding to justify the extra work.