If yesterdays conference focus was on innovation, (not just push-pin maps !!), today the second and last day of the conference focused on some of the enabling technologies of Where 2.0 – in other words, positioning technologies, GIS software tools and geospatial data.
The main story of the day I guess was the continued collective works of the Open Source Geospatial Foundation or OSGeo, which after a rather shaky start seems really to be pushing ahead, providing an environment to develop tools like MapServer, Mapbender, PostGIS etc. Indeed for me one of the unsung heroes of GIS today Frank Warmerdam, orchestrator of the GDAL toolkit used almost everywhere across the GIS industry, presented on the power of such open source libraries.
The big buzz in terms of positioning technology was Loki the consumer focused product of Skyhook, who have developed a tool that locates wifi enabled devices using the location of hotspots. To do this of course they need to know where the hotspots are and they are currently driving North America, Australia and South Africa to build their database – industrial warchalking !!!
As Nathan Torkington conference co-chair noted, much of what we see now presented at where 2.0 and on the geospatial web is dependent on the availability of data, the new community developed data as well as the more established base infrastructure datasets. I was rather disappointed with the ‘Future of the Data Industry” session, some big questions around quality, currency and general “fitness for purpose” of data were not really addressed – the future I don’t think is increasingly high resolution imagery as suggested by Microsoft, once you get to 10-15cm resolution imagery do you really need to go much higher.. what Where 3.0 applications will need is intelligent semantically rich datasets.
So my best of show along with many others I guess is MetaCarta’s geoparsing tool, which geocodes geographic terms in any web accessible document. Geoparsing is not new I remember how cool I thought it was when demonstrated to be at Edinburgh University a few years ago, but MetaCarta has built an API to their service to allow easy integration with other applications.
To demonstrate they have geoparsed a number of the texts available as part of the Gutenberg project, and produced maps of the locations involved at www.gutenkarte.org . My Favourite is the map of the locations from H.G. Wells “War of the Worlds” which was set in my local part of England, the map is not perfect, sometimes the wrong locations are identified – but you get the idea.
I need to spend a few days gathering my thought’s about the conference – I can’t remember a conference I have attended in the past few years where this was just so much to take-in.. Where 2.0 will be again taking place in San Jose next June, and will be a show not to miss.
Written and submitted from the Fairmont Hotel, San Jose, using the hotels broadband network.
3 replies on “Where 2.0 Day 2 – It’s all about data..”
I’m sometimes surprised how little awareness there is about geoparsing in the professional GI community. It allows you to turn your GI system into a full-blown geocoded search engine. All you need is a gazetteer and a geoparsing engine, though the engine needs a fair bit of intelligence e.g. natural language ability.
Geoparsing is well established in the oil industry, as it’s especially useful for quick analysis of legacy, scientific and competitive intelligence material, looking for clues where new business opportunities might lie for oil and gas exploration worldwide. We are using it to geocode entire EDMS systems with millions of documents. Linking it to scientific journal libraries is another powerful application but the publishing houses are not unlike your sleeping bears…
Not surprisingly Metacarta is a CIA spin-off as it’s well suited for intelligence work, but also has a multinational oil company as major investor. You can indeed integrate a tool like this relatively easily into your own IS landscape, though often that is already cluttered with existing GIS viewers on top of business or science applications… But you don’t always need to draw or see a map to get the benefit of spatial analysis or geoparsing.
Thanks for the write-up, Ed. Good to hear about some of the conf highlights. I checked out metacarta GeoParser API’s, look pretty nice. Wasn’t apparent from their home page where the docs were, so here’s the link for anyone who wants to peruse:
[…] response (which I’ve posted on his blog as a comment to his post) is that none of the APIs comes from UK-based companies, apart from the BBC. They’re all […]