Wednesday, August 01, 2007

Australian GeoNetwork Developers Group

Geoscience Australia BuildingOn 1 August 2007 I attended the initial meeting of the Australian GeoNetwork Developers Group at the impressive Geoscience Australia (GA) green building in Canberra. These are my notes from the meeting (not official minutes). The GA building is worth visiting, even if you are not having a meeting.

Geoscience Australia

... Geoscience Australia plays a critical role by producing first-class geoscientific information and knowledge. This can enable the government and the community to make informed decisions about the exploration of resources, the management of the environment, the safety of critical infrastructure and the resultant wellbeing of all Australians.

From: About us, Geoscience Australia, 2007

Library courtyard in the Geoscience Australia BuildingGA have a public exhibition of minerals and seismic instruments, as well as a map shop (which also sells polished rocks) and a library and cafe, which the public is welcome to use. This is well worth a visit for tourists. Also they are having an open day 26 August 2007.

Geoscience Australia provides part of the Australian Tsunami Warning System (such systems haven't had all the bugs shaken out of them yet). GA also acquire seismic and other mapping data to help with mineral exploration, natural resource use and are looking to provide access to it via the internet, which is why they were hosting the meeting.

Office of Spatial Data Management

The meeting was called by Ben Searle, General Manager, Office of Spatial Data Management in Geoscience Australia, which looks after Government mapping policy:
The role of the Office of Spatial Data Management (OSDM) is to:
  • provide administrative support to the Spatial Data Policy Executive (SDPE) and the Spatial Data Management Group (SDMG);
  • implement the workplan and manage the working groups established by SDMG;
  • facilitate sharing of experience and expertise between Australian Government agencies;
  • provide technical advice to the SDMG;
  • promote efficient use of Australian Government spatial data assets;
  • represent the Australian Government's interests in spatial data coordination and access arrangements with the States and Territories; and
  • foster the development of a private sector spatial information industry.
From: About OSDM, Office of Spatial Data Management, Geoscience Australia, 25 Jan 2006
Australian GeoNetwork Developers Group

There were about 20 people present at the meeting in the Geoscience's Scrivener Room (the room has a wavy ceiling which improves the acoustics, but the computer controlled daylight adjusting lights were distracting). This was scheduled to start at 9:30am, but I was 20 minutes late, just as Ben was finishing the introductions. Here is the agenda annotated with my notes:

Agenda

1. Introduction, Ben Searle.

2. Overview of Meeting objectives, Ben Searle: Ben suggested the need for both companies and researchers to be involved. He suggested that open source should be used and that Australia needed to work with international standards.

This what was on the agenda as Possible Meeting Objectives:
  • Establish a management mechanism - Terms of Reference?
  • Agree on a ‘single point of contact’
  • Determine who wants what and who can contribute resources
  • Establish short and long term needs
  • Determine need for a technical meeting
  • Identify the User Community
  • Identify possible software developers
  • Agree on open source principles
  • Identify possible resources
Much of what was discussed in the meeting was geospatial specific and my only experience of that was helping with the Sentinel fire tracking system. (including an experimental alternative web interface designed for mobile phones).

3. Overview of GeoNetwork application, Kate Roberts: Kate talked about the BlueNet MEST project:

The BlueNet project will establish a national distributed marine science data network linking universities to the AODC, to support the long term data curation requirements, and data access needs of Australia’s marine science researchers.

BlueNet will build infrastructure to enable the discovery, access and online integration of multi-disciplinary marine science data on a very large scale, to support current and future marine science and climate change research, ecosystem management and government decision making. ...

From: BlueNet, University of Tasmania, 2007

OCHA Maps-On-Demand

BlueNet are using the GeoNetwork open source software. Their system is up and running but most records are not yet available to the public. However, the system has a similar interface to other GeoNetwork implementations, such as the Office for the Coordination of Humanitarian Affairs' (OCHA) "Maps-On-Demand".

Kate mentioned ISO 19115, the Geographic Information Metadata standard from ISO , the Z39.50 protocol and problems with security got a mention (LDAP seen as the solution). Problems with flexibility of the software for handling XML data and IP of different data sets (including provision for Creative Commons). ePrints also was mentioned. Many of these issues were familiar, particularly how to share information while retaining the owner's rights.

Kate then gave a demo. Unfortunately the text was so small I could not read any of it, apart from the logo on the top of the screen. Perhaps UTAS needs an web accessibilty course.

One question asked was how to use the thumbnail images on the right side of the screen. At first these seemed to be purely decorative and so the issue did not seem relevant. But on the next screen it turned out that this was where on the screen commands were displayed.

The term "clone" was used to indicate "copy". This was confusing and also is potentially emotionally loaded for the general public, with the debate over human cloning.

A very complex nested metadata form was then shown. This could be useful for metadata experts (and the students I teach metadata to), but will be unusable for the average user. A simpler web search type interface is needed.

My only quibble with the technical standards is that
GeoNetwork use of Z39.50 is a bit dated (and something only a librarian could love). Web Services would be a better idea. However, Z39.50 might be needed to interoperate with other repositories.

The OakLaw project (
Open Access to Knowledge or OAKL), got a mention as they are looking at adapting Creative Commons for Australia (Prof. Brian Fitzgerald from QUT talked about it at the National Scholarly Communications Forum 2007). It was suggested that the Queensland Government was looking to use CC for government data. This then could be applied to Commonwealth data. The people doing Oak Law at QUT have already produced an Australian version of the Creative Commons license. This would seem to be adequate for use by government agencies (but I am not a lawyer).

The demo then showed a map of Tasmania, at different resolutions and pop-up windows of data from features on the map. The interface could do with some of the user friendly features of Google Maps.

In some examples the thumbnails were small maps, which looked useful.

4. Discussion on Governance Mechanisms and Related Issues, Everybody:

This what was on the agenda for Governance Mechanisms and Issues:
  • How best to arrange, manage and coordinate our activities?
  • How often do we meet?
  • Do we need a technical group to support a management group?
  • Do we need a single point of contact?
  • Who is best suited, interested and willing to perform this role?
  • Should the point of contact be funded?
  • How do we develop and coordinate specification development?
  • How do we prioritise development activities?
  • Do we allow participation of the commercial sector?
  • Can they assist in they management, development and/or as project participants
There was a discussion of what mechanism for work for government agencies. This reminded me of the administrative process I helped invent invent for the creation of Internet networks and web sites of the Australian Government. This started out with a self appointed group (the Commonwealth Internet Reference Group formed in 1994) and was later formalized. I suggested a similar strategy with OSDM as the lead agency and endorsement from AGIMO.

The
Australian Bureau of Statistics (ABS) is taking a similar approach to statistical data coordination with its National Data Network, as is proposed for geo data. I suggested using the administrative processes and terminology developed by ABS.

5. Identify Priorities, Everybody:

This what was on the agenda for Identify Priorities:
  • What are the key technical issue that need resolution first?
  • Do we need to hold a more technical meeting to commence the specification development?
  • Identification of resources including agencies willing to support the development process, funds and other resources
  • Do we need a short term and long term objective and can these be carried out concurrently?
There was then a discussion of software tools needed. Many of these were geo specific. Some were to do with XML data validation. There was discussion as to who might do software development: government agencies, companies and/or universities. I suggested that ANU students were capable of producing software, but this needed to fit in their educational program, which requires a six or twelve month cycle. It is easier to fit open source prototypes, tools and research into education, rather than proprietary production code.

Students who have undertaken ANU's "IT in e-Commerce" (COMP3410/COMP6341) and similar university courses, will be familiar with
XML, XSL, DTD, CSS, knowledge discovery, Web Services, meta-data, web-based data mining, data management, security, encryption, authentication and the like. But they will still struggle with the problems of the politics of data.

There was discussion of Wikis and mailing lists (the sort of thing used for the ACS Green IT Group is using). There was also a discussion of metadata entry tools.

There was a healthy skepticism as to the status of international standards. Curiously there was no mention of Standards Australia.

6. Initial Project Plan and Timeframes, Everybody:

To be delegated to a meeting of steering and technical committees.

7. Summary, Ben Searle.

Additional Information - Possible Areas for Discussion

Meeting closed at 12:25pm.

Further thoughts

This was a very useful meeting, with people expert in the field and from leading organizations. However, a perception that government committees need to work in a particular way seems to be hampering progress. Use could make of Web 2.0 and social networking technology for consulting and coordinating the work. In this way the inertia of conventional committees could be avoided.

A major problem with efforts such as the Australian GeoNetwork Developers Group is to find who may want to be involved. This can be overcome by placing the information online so that interested people can discover it. The next step can then be taken to invite them to comment and participate. Rather than a rigid plan, anyone interested can be invited to participate, using generally agreed standards and open source systems.

An example of where a looser method of coordination was used was in the
web based open source disaster management system for an Indonesian earthquake. Instead of conventional documentation, the Indonesian IT students doing the work convinced me that a Wiki could be used. The result was a more social, human and inclusive document than would be usual for an IT project.

In Which Repository?

The approach to metadata and repositories for geodata is much the same as that used for other types of data, such as statistics, documents and cultural records. The geoscience community have much to gain from being able to work with other such communities of interest and much to loose if they do not.

As an example, the ANU has an electronic repository Demetrius (named after the first Librarian of Alexandria). The holds mostly materials from the humanities, with culturally significant archives, such as photos of pubs of NSW. Geoscience also holds electronic copies of research publications across disciplines. If the Geoscience materials are not visible in the general repository it may never be found by potential users. Policy makers may not not even notice that geoscience is making a useful contribution and therefore not fund it.

Courses

ANU is offering courses in its System Approach to Management of Government Information. This was developed with the for the National Archives of Australia for teaching e-document management to public servants. It includes a short version of my metadata/e-repository lectures. This could be expanded to include more scientific aspects of metadata and geodata. That would be much more interesting for the students than learning how to file government paperwork electronically. ;-)

Smart Rooms

One aspect of Geoscience is the need to have computerized measurement equipment in the field. Following up on the lunch discussion after the meeting with some of the participants,
my proposal for a transportable smart room might be useful. As well as being used for school children at remote indigenous communities and for command and control on the new amphibious ships HMAS Canberra and HMAS Adelaide, the technology could be used for geosciece at remote locations, with something more modest than used for arctic research.


ps: There seems to no ISO 19115 entry in the3 Wikipedia in English. Perhaps Australia could contribute one.

No comments: