The Board Room

1. Who?

  • Implementers
  • Companies (even the big ones; esp. online services)
  • Geographers of place (work on theory of place)
  • Ontologists
  • Scientists who use place as type sites
  • Not just people working in North Atlantic!

2. Topics

  • didn't find the 3 topics very memorable; couldn't really remember what was presented in which track
  • alternative words (possibly with same problem, but anyway):
    • tools/methods
    • content
    • purpose/use
    • theory
    • interoperability is important, but ingrained in everything else

3. fundamental questions and contexts

  • NO - there is not a rigorous theory of gazetteers
  • but YES for ADL
  • the group couldn't agree that there was a rigorous theory

4. gaps/research/developments needed

  • minimum set of components
  • gazs as knowledge organization systems
  • when is a gaz not a gaz
  • role of qualitative informatoin (georef)
  • generic vs. purpose-specific
  • what would a rigorous theory be?
  • provenance/authority
  • temporarily
  • modeling footprints
  • what is a place?
  • populating
  • maintaining
  • validating
  • interoperating

Hunt Room

spent most of time on question 3

3. fundamental questions in the fieldd

  • standardization (ISO, OGC) - visceral reactions - but we didn't touch on what's going on int he formal standards community
  • ISO standards heavily weighted toward textual identifiers
  • so what is a gazetteer -
    • OED a geographical dictionary
    • ISO instance of a class or classes of features that have some positional information associated with them
    • UN ordered lists of toponyms that have typological and other information associated with them (including locational)
    • interesting thing is: evolution of defintiions over time have expanded the defintiion of gazetteer
      • in all cases, the evolution of definition expands to include more kinds of information - so by the time you get to UN, the gazetteer can include very rich information
  • multiple defintions of a gazetteer - the gap is what these working definitions (context-based), for what gazs are, depending on the focus of the community using the gazetteer
  • establishing this family of definitions may be a fruitful pursuit

1. who's not here?

  • "the users" - but there are users here ... but who are the consumers of gazetteer information
    • 80 billion people with cellphones - they don't need to be invited
    • proxies for using communities (partic. where those proxies are the providers to those communities)
  • representatives of the standards community?
  • the grid community
  • funding agencies
    • we are now at the point where they wouldn't be embarassed to come to such a thing

2. the 3-part division

  • it did help organize this meeting and successfully guiding the assembly of people here to talk about this stuff

4. gaps

  • expanding use cases
  • test environments

Carriage Room

1. What organizations, individuals, opinions are not represented?

  • Google, Yahoo, MS (not just local and mobile, but also general search indices seeking geoawareness
  • CompSci? Knowledge representation folks (only Jerry Hobbs is here)
  • Psychologists studying place awareness
  • Cognitive linguists
  • Online travel industry (expedia, hotwire, etc..)
  • small cultural projects don't realize that geography is the question, so they wouldn't really know that they should be here (another example, chowhound.com)
  • ones who know geography is the link, but don't think about how to do it the same way we do
    • extend the dewey decimal classification to cope with places
  • Opinions:
    • web tools social networking (e.g., myplace)
    • collaborative atlases - platial, 43placs, wikimapia
    • where2.0

How can they be included?

  • format of workshop limits participation- small group, hand selected to present research
    • to exapnd, it would help to have bigger/wider organization
    • ideas?
      • actual coding projects the group works on in an open source sense with associated trappings of community and release
      • climate community has communtiy-based modeling by volunteers
  • how to attract commercial entities?
    • don't know *why* they're not here
  • meet in a bigger hall - but then not so productive a meeting?
  • different models for organizing workshop/conference
  • breakout sessions earlier
  • NSF/ESF - alot of these questions are fundamental research questions
  • Census/postoffice type people

2. Are the 3 topics (components, process, interoperability) appropriate?

  • appropriate if we're trying to do what gazetteers have done in the past, but if we're trying to do new things we need new components
  • what the new components are depend on who we're bringing to the table

are there other ways of organizing the field

  • components are abstractly appropriate and not complete, tailored to a particular audience
  • services another topic?
  • collaborative tools and processes
  • need to define core of gazetteer research to put the broader stuff in perspective
  • include one topic to deal with various kinds of peripherals -- don't become a walled garden
  • eg geo info retrieval
  • comments on the fundamental organiztion and process of the sessions
  • 3 topics are interesting, adding more to the lest and choosing a different
    • collaboration (peer2peer, wiki, open formats, mashups ...)
    • "knowledge representation?" theory of locative structures = ontologies, formal presentation of relations (when is a place a place, how to deal with change)
    • use case analysis and categorization of same; tied to different user communities and application domains
    • authorities and provenance - broad track including that on trust, respect, etc.
  • need more meetings, not more people at meetings
    • other models - lightning talks, but keeping the long discussion

3. What are the fundamental questions, principles, concepts of the field

Is there a rigorous theory of gazetteers

  • there is not wide agreement that there is a single theory
  • yes, there could be one
  • but there could be formal, top-level notions that local gazetteer theories could be aligned to
  • what is the theory for if we do have one? do we want one?
  • a formal representation makes it possible for 2 different projects to find a common ground
  • theory or defintiion -- ontological way of speaking different from compsci
  • common set of vocabulary about what the theory would mean, and then pick
  • worry: something too rigorous and defined can exclude useful and appropriate things
  • definitions of gazetteers have expanded, for many of us, over the last couple of names
  • what is a name? could it be an id?
  • principals in theory oriented toward preventing exclusion and enabling contact between otherwise disparate activities
  • should we be seeking "common framework"
  • a technology stack - upper level notions that all can agree to
  • common understanding
  • there is a set of applications where wee'd like to do the same things without reproducing the same tasks and tools - this is our motivation for
  • reduce duplication in the domains of effort and code and data-amalgamation
  • not interoperability in the sense of OGC Interoperability
  • set of guiding principles - that communicates to someone what they can expect to get from being involved, and so they will know what to produce (or what not to produce)
  • current practice is not largely focused
  • database management - we all had our own files and data b/c no ability to interoperate; pragmatically, now this is the state in geodata, but maybe won't be in the future
  • a definition of gazetter: 2 element
    • a string of glyphs that are known (text), interpretable
    • source id (where that string of blyphs came from)

4. What gaps exist in our current knowledge

what research is needed

what developments are needed

  • a clearly explained list of applications of gazetteers - so they can be studied, categorized, compared, etc. (an zoology of gazetteers)
    • things called gazetteers
    • things that use gazetteers
  • coalesced semantic structure of what we're trying to talk about
    • what are the limits
    • can one gazetteer something?
  • relationship between gazetteers and more common knowledge organization systems (ontologies and so forth), which often include the kind of information we have
  • exhaustive list of types of geographic somethings that one could consider putting in something might call a gazetteer
    • MGRS coordinates
    • postal addresses
    • telephone numbers?
    • is a gazetteer a DNS system
  • the "names layer" - could we broaden to the "annotation layer"?
    • gazetteers would be part of the annotation layer
  • looking at suite of things people are
  • no tools- gaps are huge
    • for building descriptive text
    • for analyzing descriptive geographic text
    • change between 2 versions of text
    • something that looks for discrepancies and errors
    • analogy to GIS tools, but for descriptive text in gazetteers
    • comparative and analytic tools that extract
    • extract gazetteer entries from raw GIS data
  • a professional society? an open-source style community? - need a place/institution to point to?
  • trust