The ECAI Clearinghouse - a spatial metadata repository indexing 'cultural datasets' - was developed by my team at the University of Sydney from 1998, along with tools for uploading metadata and generating interactive maps on the fly using TimeMap. The system is still operational on the ECAI server, although no significant development work has been done on it for the last few years.

In many ways, the ECAI clearinghouse was a precocious Web 2.0 application, intended as a forum for sharing information, building maps from data shared by others, and publishing those maps back into the system. The technology of the time was not conducive, however, to the sort of instant, zero-barrier publishing we associate with social network sites today. The clearinghouse also suffered from the restricted size of the interested community, the heterogeneity of what was understood as a 'cultural dataset', and the lack of tangible rewards for participation.

Over the past three years, we have been developing a new web service called Heurist (HeuristScholar.org) which can be described as an academic social bookmarking database of just about anything (bibliographic records, web bookmarks, notes, annotations, historical events, people, sites, multimedia - currently more than 70 record types).

As part of the development of Heurist, we are duplicating the metadata storage capability of the ECAI clearinghouse through spatial dataset record types. Not only will this allow us to duplicate much of the functionality of the clearinghouse in a modern social web application with all the advantages of tagging, annotation and linking of records that this implies - but the increased flexibility of Heurist allows direct digitising of locations within the system and allows spatial data (notably KML files) to be attached directly to Heurist records and included in the maps generated from them. The use of alternate web mapping technologies - TimeMap, Google Maps, Google Earth and Openlayers - as well as timeline and network relationship functions, also gives users greater flexibility in generating lightweight spatio-termporal visualisations for use on the web.

This paper will include an overview of the ways in which spatial data can be stored and manipulated in Heurist, as well as specific examples of mapping applications developed with the Heurist API, Cocoon and XML feeds.