FY 2002 Innovative proposal 34010

Additional documents

TitleType
34010 Narrative Narrative

Section 1. Administrative

Proposal titleVisualization Tools for Information Discovery and Decision Support
Proposal ID34010
OrganizationPacific Northwest National Laboratory/Battelle (PNNL)
Proposal contact person or principal investigator
NamePatricia A. Medvick
Mailing addressP.O. Box 999 Richland, WA 99352
Phone / email5093752147 / [email protected]
Manager authorizing this projectJohn Calhoun
Review cycleFY 2002 Innovative
Province / SubbasinSystemwide / Systemwide
Short descriptionFacilitate decision making by providing a suite of software tools for information access and visual analysis that help to quickly find and extract specific information from a vast array of text documents (reports, journals, messages, e-mail, etc.).
Target speciesThis software would facilitate finding information relevant to a particular population
Project location
LatitudeLongitudeDescription
PNNL offices in Richland, WA
Reasonable and Prudent Alternatives (RPAs)

Sponsor-reported:

RPA

Relevant RPAs based on NMFS/BPA review:

Reviewing agencyAction #BiOp AgencyDescription

Section 2. Past accomplishments

YearAccomplishment

Section 3. Relationships to other projects

Project IDTitleDescription

Section 4. Budget for Planning and Design phase

Task-based budget
ObjectiveTaskDuration in FYsEstimated 2002 costSubcontractor
Outyear objectives-based budget
ObjectiveStarting FYEnding FYEstimated cost
Outyear budgets for Planning and Design phase

Section 5. Budget for Construction and Implementation phase

Task-based budget
ObjectiveTaskDuration in FYsEstimated 2002 costSubcontractor
1. Define ingest paths for documents a. Document labor-intensive ingest methods 0.5 $0
1. b. Explore commercial products 0.3 $0
1. c. Establish paths for document ingestion 0.5 $0
Total Cost for Task 1 $19,770
2. Define format for documents in repository a. Evaluate text chunking styles 0.3 $0
2. b. Automate chunking style 1.0 $0
Total Cost for Task 2 $30,285
3. Identy segments containing noise a. Find and document meaningless segments 0.2 $0
3. b. Automate noise removal 0.5 $0
Total Cost for Task 3 $12,942
4. Optimize interface a. Establish sets of example queries 0.3 $0
4. b. Define common query patterns 0.4 $0
4. c. Explore script creation 0.5 $0
4. d. Implement query specializations 1.0 $0
Total Cost for Task 4 $23,727
5. Integrate system for Fish and Wildlife a. Ongoing integration of implementation 1.5 $0
Total Cosr fir Task 5 $22,945
6. Test system to insure quality a. Iteratively test throughout development 1.0 $0
6. b. Conduct workshops for feedback 1.0 $0
6. c. Final version testing 1.0 $0
Total Cost for Task 6 $58,329
7. Deploy a. Establish software at Federal agencies 0.5 $0
7. b. Create repositories 0.5 $0
7. c. Provide introductory training & support 1.0 $0
Total Cost for Task 7 $13,346
Subcontractor Costs (Anadromous Fish Ecology, Columbia River Research Lab, U.S. Geological Survey, Cook, WA) (To be paid directly by BPA) $18,523
Outyear objectives-based budget
ObjectiveStarting FYEnding FYEstimated cost
Outyear budgets for Construction and Implementation phase

Section 6. Budget for Operations and Maintenance phase

Task-based budget
ObjectiveTaskDuration in FYsEstimated 2002 costSubcontractor
Outyear objectives-based budget
ObjectiveStarting FYEnding FYEstimated cost
Outyear budgets for Operations and Maintenance phase

Section 7. Budget for Monitoring and Evaluation phase

Task-based budget
ObjectiveTaskDuration in FYsEstimated 2002 costSubcontractor
Outyear objectives-based budget
ObjectiveStarting FYEnding FYEstimated cost
Outyear budgets for Monitoring and Evaluation phase

Section 8. Estimated budget summary

Itemized budget
ItemNoteFY 2002 cost
Personnel FTE: 0.7 $59,428
Fringe $15,511
Supplies $5,200
Travel 3 2-day trips to Portland for 4 PNNL personnel $9,732
Indirect $67,745
Capital 0 $0
PIT tags # of tags: 0 $0
NEPA 0 $0
Subcontractor Anadromous Fish Ecology, Columbia River Resources Division, U.S. Geological Survey, Cook, WA $18,523
Other Fee $23,728
$199,867
Total estimated budget
Total FY 2002 cost$199,867
Total FY 2002 budget request$199,867
Cost sharing
OrganizationItem or service providedAmountCash or in-kind

Reviews and recommendations

This information was not provided on the original proposals, but was generated during the review process.

Recommendation:
Fund - Rank 11
Date:
May 24, 2002

Comment:

This 11th ranked proposal is innovative and fundable. It is a good proposal to address a potentially monumental problem facing the scientific community in the region. The PI's from PNNL propose to specialize PNNL software called, IN-SPIRE, to help one quickly extract and use information from electronic documents. IN-SPIRE has been developed to accept large volumes of unformatted text, determine dominant topics and relationships within the text, and present the results in a visual format that can be interpreted intuitively. This approach has the potential to uncover hidden information in large document collections and facilitates finding pertinent information without requiring investigators to read through every document in a collection.

However, the ISRP was not totally convinced that the software is really better than use of an existing search engine plus one's own well-tuned judgment. On the one hand, information glut is a real problem, but much of the problem comes from the glut of low quality information that must be sifted and weighed with judgment. This proposed software does not address information quality.

A drawback is that apparently the software is classified and could only be used by federal natural resource management agencies. The PNNL IN-SPIRE software suite was originally developed for the U.S. intelligence community. Use of IN-SPIRE will allow federal agencies involved in Columbia River resource management to analyze textual information from a variety of sources more quickly and more comprehensively.

The staff at PNNL have the technical training and experience necessary to have a high probability of producing a useful product. A primary question that should be addressed during the contracting period is access to the software by the public given that the project will be funded through the Council's Fish and Wildlife Program. In particular, access to the products of the project by state and tribal agencies would seem to be necessary. Even if the software cannot be made available to non-federal agencies, the cost of this project seems like a bargain for providing information to a substantial proportion of the scientific community in the Northwest U.S.


Recommendation:
Recommended Action
Date:
Jun 28, 2002

Comment:


Recommendation:
Date:
Jul 12, 2002

Comment:

Statement of Potential Biological Benefit
No benefit to listed species.

Comments
There is absolutely no merit to this proposed work. There is no need for a computer based information discovery tool to support decision making on natural resource management issues in the Columbia River Basin. The premise of this proposal is flawed: there is so much information appearing in reports and publications that decision makers cannot be properly informed to make optimal decisions. While it is true that the volume of reports generated by natural resource management agencies is daunting, the information content is so negligible that it is a trivial exercise to maintain a detailed understanding of the current state of affairs. Staying current on a technical topic is the responsibility of technical staff, their activities or lack thereof, and cannot be replaced by a glorified search engine. Finally, an expert search tool will not replace quality work and content in report, publication, and proposal presentation.

Already ESA Required?
No

Biop?
No


Recommendation:
Date:
Jul 12, 2002

Comment:

Statement of Potential Biological Benefit to ESU
No benefit to listed species.

Comments
There is absolutely no merit to this proposed work. There is no need for a computer based information discovery tool to support decision making on natural resource management issues in the Columbia River basin. The premise of this proposal is flawed: there is so much information appearing in reports and publications that decision makers cannot be properly informed to make optimal decisions. While it is true that the volume of reports generated by natural resource management agencies is daunting, the information content is so negligible that it is a trivial exercise to maintain a detailed understanding of the current state of affairs. Staying current on a technical topic is the responsibility of technical staff, their activities or lack thereof, and cannot be replaced by a glorified search engine. Finally, an expert search tool will not replace quality work and content in report, publication, and proposal presentation.

Already ESA Req? No

Biop? No