1. U.S. Department of the Interior U.S. Geological Survey FGDC CWG Meeting October 10, 2006 Geospatial One-Stop What’s New? Rob Dollison rdollison@usgs.gov

2. What’s New – Current Activity  Two new part-time IPAs from State and Local Government working with GOS this year to facilitate State and Local publishing– Sam Wear and Edith Penokpa.  New community application and tools development underway at the NGPO  Agreement with Naval Research Lab’s GIDB Portal for publishing additional web services to GOS

3. What’s New – Current Activity  Marketplace grew from 500 plans to over 2,000 in FY06  Overall Collection grew from 81,000 to 117,000 records in FY06 (30% increase)  Portal now funded for three new enhancement release cycles in FY07  Stakeholder Design Team Activity underway

4. Issues Recorded in our tracking system

5. Improvement Schedule AudienceImprovement Oct-Feb Build 2.1 Mar-Jun Build 2.2 Jul-Oct Build 2.3 Users  Search  Marketplace  Viewer and Base Maps Research and Testing TBD PublishersHarvesting, Reporting, Feedback TBD OMB and Agencies Reporting Research and Testing TBD DevelopersUDDI Research and Testing TBD

6. Search Enhancement Activities  Improve Spatial Relevance  Standard versus ‘Advanced Options’  Text only vs. Spatial  Additional Advanced Boolean Capabilities  Export Results to Spreadsheet  Possible view results on map  Metacarta – evaluate for possible text based collections – Grants.gov and possible other applications

7. Ranking the Relevance of a Spatial Match Given a query polygon Q, figure 1 shows how we score the spatial relevance of target information object T. The intersection, T ∩ Q, is called X. If the information in T is uniformly distributed, then the fraction of the information that is contained in X is F t, where F t = X / T. where 0 ≤ F t ≤ 1 (Equation 2) and X, T are the areas of X and T, respectively. With similar reasoning, we can presume that, since X only covers part of Q, only a fraction of our information request is fulfilled by X, or F q = X / Q where 0 ≤ F q ≤ 1 (Equation 3) and Q is the area of Q. Using area to compute the fractions assumes a uniform distribution of information within T and a uniform importance of information throughout Q. Using non-uniform distributions would require more computation, but the principle is the same. A composite spatial score, S, will be the product of the two fractions, S = F t F q. (Equation 4) The value of S will range from 0 for no match to 1 for a perfect fit. We can focus the power of this test, increasing the penalty of mismatches, by raising the fractions to powers k t and k q for target and query, respectively. The spatial score than becomes, S = F tkt F qkq if F t, F q > 0 (Equation 5) S = 0 otherwise. Increasing k t raises the importance of the target being entirely with the query area, and decreasing k t indicates a willingness to accept targets that extend beyond the query area. Increasing k q raises the importance of finding targets that cover the whole query area, and decreasing k q indicates a willingness to accept smaller targets within the query area. Setting either to 0 results in treating that part of the test as a Boolean search. QTX Figure 1. Diagram of query polygon Q intersecting target object T. The intersection area is X. A Spatial Overlay Ranking Method * From Ken Lanfear paper

8. Reporting Improvement  Better feedback and reports to our publishers on their metadata and data use  Make publishing easier – ‘download’ in metadata will have a download button  Improve QC – Harvesting validation rules and ‘get map’ service checker  Support FGDC Annual Report and other reporting requirements

9. Marketplace Improvements  Currently harvesting NDOP, NDEP, and Ramona plans  Add Search filter by data type? e.g. Imagery, Lidar, etc to better visualize plans  Highlight email notification options – receive emails about new plans in your area  Export results to spreadsheet  Possibly add discussion capability to marketplace

10. Community on Home Page

12. Community Application Development Tool Developed by Marc Levine (NGPO) and his team – a tool to develop an application from live services with out having to be a ‘developer’ Draft Output Example: http://mapdss2.er.usgs.gov/hazards/map s/