sotm 2008 – keynote

Robert Bar — KeyNote

Social Geographer since the early 1970s
interested in where people live, how they live and where they worked?
we seemed to know less about this then the history colleagues as we were relying on old census data

in the early 80s we realised that the data had to be liberated as most data then had a address related to it

address is an textual street reference system

Gur Mukh Singh, used the opportunity of a recession and went into remote areas to setup map making factories. But when the first data set came out had significant errors and this was not going to prove useful at all. He was waiting for people to buy the data before producing it. It had a agreement with OS that they would not compete with him.

Christopher Roper and Venssa Lawrence setup a company to produce data sets for sale. The first project was to produce a data set called NORA….

Roper moved to OS and then started work on producing address point data …

Speaker thinks that Navteq, Tele Atlas and Google are headed towards building a pay per use system for street addresses. Such information is of the public and should belong to public and not to some Mega Galatic company.

There are multiple problems with collecting GeoData. Topology, exact location, correct names and correct spellings are really important, but shape is not so. Topology and attribution are important to developing navigation solutions.

Companies like ITIS are able to provide real time traffic feeds using 2 way GPS devices on fleets of cars, anonymising their data and selling it on to third parties. Yes GPS feeds are dirty, but collating many dirty feeds and analysing them can result in clean data feeds using statistics. This should provide a faster way of mapping the country and the world then sending a bunch of people out and about with GPS devices during mapping parties.