The infrastructure session considered what technical services might be necessary to facilitate the analysis and synthesis of data with a spatio-temporal component. These might be APIs that customisable tools might use or simply human-readable functionality for querying or visualization. Specific requests for research and investment included:
- Vernacular mapping. Although many web services are now available for dealing with semi-official place names, there is still the inability to recognise slang or informal references, such as the ‘Weisswurst Equator’ (between North and South Germany) or ‘The Big Smoke’ (London). These are particularly useful for the Humanities where formal references cannot be expected. Some experiments have been undertaken by mapping aggregate search results but they can be heavily influenced by ranking algorithms
- The problems of dealing with uncertainty continue to plague this topic. places and periods are vague, socially defined constructs and source data almost always leads to a imprecise and/or inaccurate data. Conventions for representing both of these aspects would be helpful, as would the ability to switch rapidly between different visualization techniques so that the ‘blind spots’ of each aren’t ignored.
- Additional ontology integration was cited as area for investment, especially between gazetteers. Data sources such as Europeana were also flagged up as being increasingly usful as a central spine for cultural heritage concepts