A while back Anthony Townsend sent me a funny blog link which had him wearing a t-shirt that said “Every time you say Web 3.0 a start up dies ”

This quickly became a running joke in the office since we had a Web 4.0 milestone running in Trac for a while. So we got a big kick out of a call a few days ago where someone referred to what we were doing as Web 3.0. The last time I’d really read anything on Web 3.0 was when the NYT wrote an article about it that bloggers had a bit of a field day with.

I figured I would take another look into it since we’d been labeled. Going to the Web 2.0 well Wikipedia kicks up:

“Web 3.0 is a term that has been coined to describe the evolution of Web usage and interaction that includes transforming the Web into a database, a move towards making content accessible by multiple non-browser applications, the leveraging of artificial intelligence technologies and the Semantic web and three dimensional interaction and collaboration.”

Lots of articles wax poetic on the issue and conflate it with the Semantic Web as in the Wikipedia definition. The Semantic Web has been around since 1999 or so and is most often associated with the thoughts of Tim Berners Lee. I’d done some research on semantic kind of things back in school and to be honest was put off by the general over complexity of it. Any time that core words to describe your work include things like semantics, ontology, lexicon etc. you are not exactly dabbling in the world of simplicity. Having spent a good chunk of my life in academia I can safely say we do an awesome job of taking simple concepts and making it so that 99% of world has no idea what we are talking about. Yes – post modernists – I’m talking about you.

My take is that simplicity forms the roots of what has made Web 2.0 successful. The API’s and defacto standards that have really taken off have the common theme of being mind numbingly simple. So, there seems to be a bit of a disconnect with Web 3.0 and conflating it with the 8 years of academic and standards work that have gone with the semantic web, which have created some very complicated white papers and manifestations.

The irony with getting labeled Web 3.0 is that what we were describing, at the time, was our attempt to simplify the world of geospatial data so that it could be consumable by non-technical people. To add to the irony there is a whole science of applying semantic web concepts to geospatial data and it is definitely not simple. Traditionally geospatial data comes in a variety of shapes and sizes – point, polygons, polylines, raster formats (satelite imagery, heat maps) etc. Part of the art to geographic science is knowing what geometries to use when – cenus tracts, census blocks, counties, zip codes etc. While this frame of thought matches up well with data formats it does not match up well with the way most people think. People think about locations and attributes or contexts about that location. I live in the Clarendon neighborhood and I associate contexts with that neighborhood like restaurants, parking, crime, housing prices, music, congestion etc. The data that describes those attributes could be a dozen different geometries, but as a user I don’t really care. I care about getting an answer to my question in the context of the location I care about – in this case Clarendon. We’ve been working on an architecture that will provide such a simplification and that along with the various other aspects we’ve been tying in is what created, at least one, Web 3.0 label. Whether what we are doing is Web 3.0 or not I really have no clue – we are hoping it solves a problem in a simple way for a user. At the end of the day that is what I think will be successful whether you label it 2.0, 3.0 or even, ack, 4.0. What is created needs to be easy and simple not only for the users but for the developers implementing it. While the next evolution will likely solve some of the problems targeted by the semantic web I think the actually technological path will be something far simpler than what is currently being touted.

***All ideas about the new architecture and contexts came from Mookie – a.k.a. Pramakta Kumar one of our lead developers. I simply regurgitate them in some semblance of an idea. The F1 platform for it all is a Chris Ingrassia creation TM.

Popularity: 12% [?]

Just thought I’d drop a quick line on the blog to let anyone who’s interested know that several of us are out in San Diego at O’Reilly’s Emerging Technology Conference.

Among the many reasons we’re out here is to get exposed to new ideas and technologies, so if you’re reading this blog and happen to be out here as well and would like to talk shop, or just go out and grab a beer to see how many bad heatmap jokes we can come up with, we’re definitely up for it.

Just drop me a line directly or add a comment to this post and we’d be thrilled to meet you.

Technorati Tags: etech, etech07

Popularity: 7% [?]

We had a request for a map of VC investments in Boston from the Rob Finn data we posted up and could not pass up the opportunity to make more heat maps.

Some interesting patterns going on in Boston with a good chunk of activity happening in the central business district and Route 128 , which has been the classic case study of regional economic advantage ( see: Regional Advantage: Culture and Competition in Silicon Valley and Route 128 by AnnaLee Saxenian ). The start ups are not stopping at 128 and you can see the dispersed activity in Framingham, Westborough, and even out to Lowell.

Always happy to fill requests, but we are burning the midnight oil to get this out of beta so you can fill your own requests. We’ve been getting some great feedback and have a cool new user interface being built out that will make it super simple to make your own map with a new world of geospatial data (at least new outside the stuffy confines of GIS). Also a kick ass new architecture to make the sometimes byzantine world of geospatial data as easy as local search.

Popularity: 7% [?]

A little while back I got a call from Rob Finn at Edison Ventures asking about our little start up and some help with a side project he was working on with his blog – VentureBlogalist.  Rob had collected a list of venture deals for 2006 by the location of the funded company and wanted to create a heat map of it.  He was nice enough to send along the list and I figured it would be a good chance to see how hard it was to morph that spread sheet into geospatial data and put it on a map.  There were 1082 funded companies on the list so I started testing geocoders to see what would get the job done.

I took my first shot with Batchgeocode.com which has a straight forward web interface that allows you to cut and paste a tab delimited list of addresses and have that geocoded and exported as KML.  Sweet – just what I needed, so I saved the excel file as a .txt tab delimited file, opened it in notepad, cut and pasted into the box and I was off and running.  Clicked the button and hit a road block – “recommend you do not geocode more than 500 addresses at a time”.  It did not work to break the data set up into chunks of 500 and I had over a thousand addresses so I hit the “run it anyways” choice.  The app auto picked the street, city, zip and state fields and all I had to do was pick the name and description fields – good to go.  So, I set the batch geocoder running and waited to see if it would handle all 1082 addresses.  First thing I noticed was my computer – a beastly Alienware laptop – was really straining and all my Firefox windows were barely functional.  About two hours later it completed, and the bad news is the KML file was empty and not usable.  Not bashing Batchgeocode.com just don’t excede the 500 limit because it no workie.  Otherwise I though the process was pretty simple and intuitive.

Bad news is I still did not have any KML to show Rob pretty heat maps with.  So I went back to the drawing board and downloaded the Juice Analytics gecoder .  Once I enabled macros in excel it was not too bad of a process.  You have to get an application ID for the Yahoo! geocoder and have a YahooID, but otherwise was pretty smooth sailing.  It is a bit of work to cut and paste your street address information and you can usually only get 100 or so addresses geocoded in one stretch before you loose connection to the Yahoo! server (do not hit debug on the macro script just hit end and highlight another chunk of addresses to geocode).

Once I got through my list I clicked on “launch in Google Earth” then right clicked on the data set and saved it as KML.  From there I uploaded it into GeoCommons added it to my workspace and made some pretty heatmaps to send back to Rob.  So if you want to see where the hottest VC activity is just look below:

I got a nice heat map of all the venture deals (green squares) Rob had collected across the USA.  The obvious hotspots showed up in the Bay Area and the Boston Washington corridor.  The activity in Denver is a bit new from previous tech spurts, but it amazes me how much things stay the same tracking innovation and diffusion of new technology.  Back in school we did lots of research on this and the trends are amazingly similar over time.  Richard Florida, now at GMU, found strikingly similar patterns of VC investment back in the eighties.  During the last Internet boom Matt Zook found the largest explanatory factor in the aggregation of domain name registration by geography was the presence of venture capital in the region – the maps looked the same as above.  In general it is a fascinating topic but I’ll not bore everyone to death getting to academic on it all.  Usually best to stick with the pretty pictures.  So, here are some close ups of the hotspots:

The usual suspects in the Bay Area – Silicon Valley innovation is not going any where

A cool close up of where the funded startups are in downtown San Francisco

Boston is still taking the lead along the BoWash Corridor, but the DC metro region is making a strong showing

While the venture hotspots was fun stuff the whole process illustrated some fundamental need to have easier tools to get unstructured geospatial data onto maps.  We’ve been working on supporting .csv and .txt uploads into GeoCommons with geocoding support, but todays experiment gave me several ideas about how the process could be a whole lot more simple.  The potential of tapping into all the data sitting in spread sheets that has street address data, zipcodes, county, or states and turning that into information that is easy to map, share and consume gets us excited for the upcoming releases.  We’ll see how easy and fluid we can make.

Popularity: 9% [?]