Aggregating The Long Tail: Sharing The Local Mojo

March 29, 2007

**Note: I’m not entirely happy with this article as written, but I had already published it with a misplaced save attempt.  I will leave it up but expect that within a few days, it will be fleshed out better and refined it a bit more (especially with any comments added)

In our first article about aggregating the long tail, we discussed the lack of spatial context with respect to closely located local information. This second article covers the sharing of local and/or micro content in real estate such that it can be successfully aggregated.

Extending The Concept of A Mashup Beyond Maps

This time we’ll start with the concept of combining information from different sources. The term “mashup” seems to have taken on a very limiting definition online.  When most people think about online mashups, they think only in terms of online maps (which require a return to the map to view other data “spatially”).  Why limit your thinking about mashups to such limiting interfaces?  The screenshot presented in the first article is actually a video mashup based off of geocoordinates, isnt it?

Given that we can easily expand our interfaces beyond online maps and stay within the definition of mashup, let’s turn our thinking towards micro-content. In the first article, we talked in non-technical terms about how to associate geolocation data around micro-content. Now we need a way to structure microcontent (about something specific – church, house, school, restaurant, etc.) and local content (about the general area) so that it can be shared and used by others  in imaginative ways…

Issue #2: Local Information Isn’t Aggregated In Any Spatially Relevant Way

Using current search technology, finding spatially related local or micro information requires either prior knowledge of an area or a third party mapping interface.  The analogy would be that if you are looking for running shoes, you wont be exposed to cleats unless you have a separate reference document to tell you that cleats are related to running shoes in some way.

Doesn’t seem like a problem until one realizes that the most relevant content isnt always displayed on the first page of  search engine result. The most relevant content with the most indicators that the particular search uses for relevancy relative to other results are.  The relevant content for you with regards with geographically co-located microcontent has a good probability of not often being found even if it does exist.  So, when I look at the first results on Google for “Birmingham Michigan”, there is some purely commercial content (they do SEO too, right), some organizational information, and some local content.  However, there is no relationship at all among this data so search engines results require some switching between various sites, entering different towns to see different results, and nothing that tells me that one local may actually be quite a distance from another…despite being located in the same city.

Additionally, data from nearby surrounding areas is only included with results that contain “birmingham michigan”

Possible Solution

A possible solution might be in having a way to aggregate and publish subsets of related local and micro-  information and submitted by users in a variety of different formats that could be consumed by the end user.  The output could be custom RSS,  another XML format, or some sort of recordset.  

Somewhat Technical Discussion

We should have some sort of formalized request for content sent to a web interface. The request could consist of as little as the following: type of content, a location, distance, and distance units (it’s my blog so let’s start out thinking about international users from the beginning here 😛 ). A time period might also be useful to filter results correctly but isnt absolutely required.

The resultant output would be a formatted grouping of the local and micro-content that meets the criteria specified in the request suitable for formatting in any type of user interface (not just maps). If local or microcontent exists, the consuming site can format or mashed as needed as defined above (using HTML, plug-in, widget, etc.)  and the result can be presented to the user.

If no content exists in the specified area, the output might represent suggestions for locations close by that have not been asked for by name but are related spatially. We use Birmingham MI as our sample content in the first article, so lets continue with that theme. So, in the case of Birmingham, MI, it might well include results from Troy, Southfield, and Bingham Farms without me knowing to for them or having to return to a different interface.

We’ll post a example of such an interface (we have a working one with our own actual data but we’d prefer to let people beat up on a reference implementation that we’ll make public over the next few days)


One comment

  1. […] – Increase discoverability of content: The rewards of great position in search engie result pages is tangible. However, there is often great non-SEO optimized and applicable short tail content that exists deep in the results of online searches. Social functionality should play a role in making such content more discoverable and assist in aggregating long tail content. […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: