I have a post up at the research e-Labs site on some of the stuff I looked at in my summer gig for the enemy, open government data and semantic publishing. The thrust:
Today’s web is largely designed for human consumption, with information too unstructured and context-dependent to be understood by machines. Consequently, any interesting manipulation of information must also be done by humans. By laying out information in a way so obvious that even a computer can understand it, we make it possible for computers to do a lot of the grunt-work now done by humans. This economises on human effort and also opens up new possibilities for the combination of data from different sources – mashups. This sort of semantic publishing – making information available in a format which explicitly flags the meaning of each piece of information – has the potential to improve the accessibility and reusability of Government information. This provides obvious benefits to the public, and also allows more efficient inter- and intra-agency use of information.
The current approach to making Government information more accessible is focused around making websites more user-friendly: the agency thinks about how people use their site and attempts to organise information appropriately. This approach is costly and there is a better alternative. Instead of attempting to be all things to all people, Government can open its information for reuse by consumers and third parties. Allowing information to be used in various ways frees Government of the responsibility of anticipating diverse and ever-changing consumer needs. Rather than providing a finished product to the public in a user-friendly form, open semantic publishing provides the raw material in a developer- and mashup-friendly form.