Wiki Navigation

Reports from breakout sessions II

wasn't able to capture the distinction during the presentation between long and short

enumerating/characterizing gaz requirements
footprints
ontology
taxonomy of users and requirements (use cases)
basic elements/mindsets of gazetteer
taxonomy of gazetteers
psych research on human categorization - cross-cultural, cross-lingual, political (medium)
community repositories and wikis
multilinguality (short term?)

community tool repository and wiki
charettes - focused sessions involving designers and users
- look at a use case, work through it, how to analyze, what possible solutions
- actually work it out in some level of detail
conferences
- COSIT
- UN conference on standardization of geonames (August 2007)

Clarifying role of gazetteers and knowledge organization systems
Vision, evolutionary direction
"Global Description Service" (universality)
Methodology for data acquistion
Concept strata
Scale-dependent behaviors of functionalities
- DNS-like model might be operated on a scale
Liability issues and IPR if wiki contexts are engaged

local protocol-free forms - freedom at local level is important - free forms for local knowledge leading to web-specified interfaces
analysis of existing and emerging stds that lead to def. or articulation points amongth them - what are the common elements
how can gaz's ....
ISO blueprints for gaz expansion of scope

forum for discusion and excahgnge
interop experiments involving alternative local designs
opengaz.org, .net, opengazetteer.org, .net - already names created
gaz. tracks in geospatial and library conferences (and other professional conferences)
build professional credential processes

1. What are priority topics for long-term basic research?

spatial natural language is vague; how to automate understanding of it, including the role of context
- engage people in natural language processing area
- looking beyond the word level to ambiguous meanings
- tools and algorithms for parsing and extracting semantic relationships
- bounding and processing spatial phrases - moving beyond entity identifications
the nature of footprints
- fundamental geometric issues (accuracy) - what is appropriate to reference, and how
- how defined
- how deal with topologically referenced footprints
support for multilinguality
- user languages - multilingual gazetteers - multilingual access to geographic data or data leveraged via language
- this is a problem of 2 parts
  - alternate names and spellings
  - scripts and transliterations
time
data acquisition and integration
how to enrich the gazetteer, esp. the type thesaurus (the knowledge organization structure) so we can capture and related to other types of information, including relationships and object-intrinsic properties (e.g., county objects and attributes for recording their area)

2. What are priority topics for short-term research?

getting real gazetteers for real people to do real things with
aggregating search results from multiple gazetteers - methods
real specific use cases, tied to test cases
what is purpose specific and what is generic - most discussion today has been purpose-specific
- how do we get to the generic? - better answered from experience
inventory of exisitng gazetteers in computer accesible form
- purpose
- content
- is there limited overlap
- we don't understand what's there
conference idea: real-life examples of gazetteer use
- here's what we're doing
- problems /solution
- here are the gazetteres we're hitting (real data)
- objective: are there common tasks and methods - what are they - core functional requirements of gazetteer services(?)
creation of gazetteers
- cheaper ways to populate gazetteers
- obsolesence
tools
fundamentals, both on data and interface sides, are not jelled
- need to find common ground where both public and private sector interests intersect
- lots of beneficiaries
community-based, social models for data acquisitiion and refinement
low-hanging fruit: produce gazetteer entries easily out of existing GIS data that already exists - as long as those records have a name
- what are the principles for storing data in multiple ways - what are the boundaries there?

3. What do we need to say to the funding agencies?

government
research
commercial
incentives more consistent for getting behaviors than policies
assumption is this will be paid for by government?
what are ways to make these things worthwhile to get them paid for in other ways than govt funding
motivation in terms of major beneficiaries
major emphasis: what is the point of doing this - what are likely benefits
there is a public requirement for efficient access to geographically referenced information
- this spans commercial, academic and governmental interests
long-term engagement is needed
deliverables to private sector need to engage and fuel them for further dialog

4. What is the future of gazetteer standards, protocols?

why we need those standards? geographic information retrieval systems on the web capable of recognizing these names
- how will this information be maintained
rephrase question: what will drive the adoption of gazetteer standards and protocols?
- there's a strong motivation for something common to emerge
  - interoperability
  - an all-knowning future thing
- emergence of a dominating gazetteer database that everybody wants to use - critical mass is essential
  - but how do you keep it current
- a model: used book sellers who advertise their books on about 3 websites
- another: sabre and the other one
- microprocessor world for getting specs out

5. What activities would continue and strengthen the momentum of this meeting, help to build an effective research community?

funding
charges for working groups
- surely there are small consortia within this community that would be interested in doing useful things
figure out who cares about these activities and why do they care?
- business/mission/use cases will flow from these
- grounding this work in real examples is important
an effective research community (a bit more inclusive) requires deliverables that are clear, marketable and palateable to those organizations/communities that we want to be part of this
- a SIG?
- more frequent meetings?
focus on finding a common purpose: a gazetteer that we could all benefit from
- is the common ground closer to the data or the techniques and approaches
  - try to find a common data set - a killer gazetteer
google earth (and the like) is perceived by a large and growing number of people (decision makers) as the earth reference system for information retrieval
- involvement of google and the like, or use of the interface, would help
- identify their agenda and peer with them