Executive Summary
Advisory Board Meeting
Harvesting Information to Sustain Our Forests
November 8, 2000
Portland Oregon
Board Members:
Attending: Craig Palmer (Chair), Monty Knudsen, Cynthia Miner, Regina Rochefort and Mark Whiting
Not able to attend: Michel Biezunski, Jeff Burley, Martin Goebel, Paul Gorman, Fred Johnson, and Steve Solem
Project Team:
Attending: Tim Tolle, Lois Delcambre, Eric Landis, Dave Maier, Craig Palmer, Patty Toccalino, Shawn Bowers and Mat Weaver
Not able to attend: Fred Phillips
Summary
Lois Delcambre welcomed the Advisory Board members to the meeting and thanked them for their willingness to assist with the project. She noted that the official starting date of the project was August 1, 2000 with funding and support provided by the National Science Foundation and cooperating Federal agencies including the USDA Forest Service (Region 6 and PNW Research Station), Bureau of Land Management and Fish and Wildlife Service. Project duration is 3 years with an overall budget of $1.5 million.
Craig Palmer reviewed the objectives of the meeting. He suggested that three purposes of the advisory board are to serve as a sounding board for project ideas, to provide suggestions to the project team, and to assist as a resource in identifying emerging technologies, standards, and test cases.
Eric Landis reviewed the motivation behind the development of the project. He pointed out the difficulties that land managers have in obtaining the information needed for wise land management decision making. The challenge of this project will be to provide tools to assist in that flow of information.
The following diagram presents an overview of the major activities in the project. Team members provided an overview of early accomplishments in the first four activity areas.
1. Understanding client requirements
Tim Tolle provided an overview of interviews that have been held with participants from three adaptive management areas (Snoqualmie Pass, Central Cascades and Applegate). Individuals interviewed included federal land managers, scientists and the public.
2. Requirements for the system
Tim highlighted the following findings from the interviews:
· People are concerned with PLACE. They would like to be able to find information (documents, monitoring results, maps, etc) about a given place.
· They seemed more interested in summary information than raw data.
· They want to be able to find study results from places that are similar to their own. They want to be able to intelligently extrapolate findings from these places.
· Most of the information they need is not in readily available in electronic format.
· A primary focus of our project should be meeting the needs of the land managers.
3. Conceptual Development
Lois Delcambre noted that the main sources of information to the system are documents. These documents can be of various types: reports, assessments, decisions, scientific data sets, environmental impact statements, etc. We need to provide a useful way to access these documents. One such method is to use place-oriented similarity searches. A place can vary in scope (a timber tract, Little Salmon River, Applegate Watershed, etc.) and places can contain other places. By attaching metadata to places – using various superimposed controlled vocabularies – we can search for similarity between places. Such similarity is defined using the controlled vocabularies. For example, a user may want to find places that have a similar climate. We can use the climate controlled vocabulary to compute similarity between places. Each controlled vocabulary (i.e. climate, topography, hydrology, etc.) will provide the framework for comparing places – using information extracted from documents.
We have defined three main entry points into the similarity search:
· GIS: A user may select a place by clicking or drawing on a map. The system would then use a digital gazetteer to derive place names from the user's selection. These place names would then be used in the similarity search to find similar places.
· Explicitly Place Selection: A user may select a place either by directly typing a place name, or by browsing a list of available place names.
· Browse Controlled Vocabulary: Instead of starting with a place, the user may search directly on features described by the controlled vocabularies. For example, “Find me documents about places with climate X and topography Y”. The system would then go directly into the controlled vocabulary to find matching places.
4. Architecture
David Maier noted that, as much as possible, we would like to use existing techniques as building blocks for building the system. We will begin by investigating a publication server application that provides a simple web interface for contributing documents (including metadata). We have permission to use the source code of this application and will use it as a starting point to get documents somewhere where they can be accessed. We have looked at other digital library systems (i.e. NCSTRL) and protocols as well.
We need some mechanism to identify documents that participate in the system. A mechanism called Digital Object Identifiers (mainly developed by the publishing industry) appears to be a useful approach. The DOI approach has several useful benefits:
· Identifies content, not location (one document could have multiple locations)
· Different levels of granularity (can address sub-document objects)
· Can address non-digital objects (instead of point the user to the object itself, it can tell the user where to go to find the non-digital object)
A digital gazetteer (produced at UC Santa Barbara) is a tool that stores geographic names and corresponding locations (points, coordinates, etc.). This tool will be very useful in relating documents to places and in evaluating user selections.
Lois Delcambre pointed out that several other ongoing projects at the Oregon Graduate institute have potential of being of use to this project.
Mat Weaver gave a demo of SLIMPad – a superimposed application.
More information can be found at: http://www.cse.ogi.edu/footprints/
Shawn Bowers gave a presentation about the architecture of superimposed information and mapping between superimposed data models.
More information can be found at: http://www.cse.ogi.edu/footprints/
Advisory Board Comments
· Board members agreed with an initial emphasis on meeting the needs of the land managers.
· A focus on “place” as a method for organizing information was also supported. However, additional methods of organizing information such as by author or by species should be investigated.
· Another focus should be on ease of use of any developed system.
· If a report or data cannot be put online, then some metadata should be created to point to that information through the internet.
· The initial information to focus on gathering is that produced by the Adaptive Management Areas themselves. Once this is collected, then the project could expand outwards to other sources of information.
· Managers should be provided with access to models in addition to information. An example is a model that provides a visual projection of what a forest will look like over time.
· Team should examine ongoing efforts to organize natural resource information by federal agencies including the Natural Resource Information System (Forest Service) and the Natural Resource Database System (National Park Service).
Advisory Board Offers of Assistance
· Monty Knudsen will provide a summary of a recent review by the National Archives of how the Fish & Wildlife Service (FWS) catalogues their documents. He will send a copy of an article in Bioscience describing the use of a virtual library to catalogue insect specimens. He also offered to provide information of the development of a document tracking and retrieval system that is tied to geographic information systems for use by FWS field offices.
· Mark Whiting offered to provide a contact at the Pacific Northwest Lab (Steve B.) with experience in associating different types of data (climate, vegetation, soils, groundwater etc.).
· Cindy Miner will send some examples of organizational themes that are not “place based” that may need to be accommodated in the overall project design.
· Regina Rochefort offered to host the next advisory board meeting in May.
Conclusion
Lois Delcambre again thanked the advisory board for their participation. She indicated that she would appreciate any additional suggestions that may come to mind. In particular, she would appreciate suggestions for test cases as the project develops. As the approach of the project is to use currently available technologies as much as possible, she would appreciate suggestions regarding individuals to contact with ongoing activities in digital libraries, electronic gazetteers, standards, or any other technologies that might be applicable to our project.
4