Although we can use NetWeaver to answer questions about CBNRM using the knowledge base we constructed earlier, information with a geographic component is often more easily understood, displayed and analyzed when presented in map format. GeoNetWeaver is a spatial analysis and display extension of NetWeaver based on the open source software MapWindow GIS (MapWindow.org).

In this tutorial we will setup a GeoNetWeaver project (here after referred to as a “geoproject”) from an existing knowledge base (CBNRM), a shapefile¹⁾, and a set of data sources; and then take a quick tour through using GeoNetWeaver. The data for this exercise are at three different geographic scales: the mythical county of Zamegal, its three regions (West, Center, East), and each of their many communities. The GeoNetWeaver project will tie together the data from these three different geographic scales and generate maps for each of the topics in the CBNRM knowledge base.

Before we can put the GIS tools to work we must set up a geoproject (GeoNetWeaver project). The geoproject defines the GIS data to use, the knowledge base, and their relationships. Open GeoNetWeaver and from the file menu, select New GeoProject… to create a new, empty geoproject. This will automatically open the geoproject setup window (figure 1) for configuring the geoproject.

The geoproject setup window is a typical wizard-based utility where you are guided, step by step, through the configuration process. As you navigate from one page to the next, the setup utility will validate your inputs and only allow you to move to the next step when you have sufficiently completed the current step.

The first step is to assign a knowledge base to the geoproject. Click the Open button and navigate to and select the CBNRM knowledge base. Once the knowledge base is loaded you should see some summary information about the knowledge base in the Info group box (figure 2).

Click the Next button.

The results page is where you specify the topics you want to map. On the left is Topics of Interest with a tab for dependency networks and a tab for calculated data links. The two lists are separate because the two types of networks are mapped differently.

We want to map all of the dependency networks. You can checkmark each one independently, or you can right click in the list and select Check All. Only the topics checkmarked will be included in the maps.

On the right is a legend editor. The legend editor gives you individual control the color scheme and labeling for each topic. We will use the default settings for now. We want the map legend to have an outline of the knowledge base with CBNRM INITIATION at its root. Click on the topic CBNRM INITIATION and checkmark “Include in outline root” in “Options” (figure 3).

Click the Next button.

Setting up the data sources is the most complicated part of setting up the geoproject. Here you will need to configure from where each data link is getting its data and set a master shapefile²⁾ and join³⁾ it with any other database tables. This exercise uses data from three different geographic scales, other geoprojects with simpler geography may contain as few as one.

Click on the database configuration button to open the database configuration utility (figure 4). Alternatively you can load a database configuration that you created previously with NetWeaver Developer.

We need to include the database tables Countries, Regions, and Communities from CBNRM.mdb (most of our needed data) and Comm.shp (our shapefile needed for its geometry and to control the other database tables).

Click on the add database tables button to open the add database tables window. From the File DBs menu select Access (*.mdb)… and navigate to and select CBNRM.mdb. Dismiss the login window as the database is not secured. Add each of CBNRM.mdb's tables by selecting each one in turn and clicking the add table button for each (figure 5). Close the add database tables window.

Now add the shapefile Comm.shp. Click on the add database tables button to open the add database tables window. From the File DBs menu select dBase (*.dbf)… (a shapefile stores its data in a .dbf file) and navigate to and select Comm.dbf (or Comm.shp, either will work here). Dismiss the login window again. This time, since there is only one table, you will be returned directly to the database connection window. The database connection window should now look like figure 6.

We need to establish one table to control all the others. We do this with the following joins:

From the Database Table Joins toolbar click on the add database table join button to open the database table join editor (figure 7). Using the drop down menus work from top to bottom in the editor to build the first join for Communities (figure 8). Click the accept button to save the join. Repeat this process for each of the remaining joins in the above table (figures 8-10). When complete the database connection window should look like figure 11.

The ID field is the field that contains the name or some other identifier for each record in the database table. While not strictly necessary, setting an ID field for each table is recommended as it lets NetWeaver know what to call each row in the table. This feature is used extensively in the automated documentation utility.

For each database table in the table below, select the table and then from the identify menu select the corresponding field. The database connection window should now look like figure 12.

Move to the Links to Data tab in the database connection window (figure 13). Using the table below set the database table and field for each data link. Once complete the connection window should look like figure 14. It' not as hard as it looks. As you select the table NetWeaver will attempt a best match for the field. All of these, because the field names are so similar to the data link names, should be automatically chosen.

If you have one shapefile and it is already set to be the master of all the other data sources through joins then NetWeaver will choose it for you. If not it is likely you missed something when doing the database joins. If you need to select it yourself choose it from the Master Shapefile drop down menu

Click the next button to move to the next step.

Everything on this page is optional. However, it is nice if GeoNetWeaver starts up with the main map displayed, so from the Show Result drop down menu choose CBNRM INITIATION.

Everything on this page is also optional. Here is a list of options including the display name of the geoproject and which tabs you want to expose to the user. For example you may not want to expose the raw data from the databases to the user. Set these as you see fit.

This optional legend is ancillary to the results legend that was configured on the Results page. It will generally contain extra map layers to supplement understanding of the maps and aid positioning. GeoNetWeaver directly uses MapWindow GIS projects as legends, so all setup and editing of the legend is done in MapWindow GIS (it's included with GeoNetWeaver).

Click the open button and navigate to and open ZamWeaver.mwprj.

When you close the setup window GeoNetWeaver calculates the results for all of the topics you selected and displays the start up topic if you chose one. Your GeoNetWeaver window show look something like figure 17. CBNRM INITIATION is displayed as a map in the left pane. On the right is a dashboard and legend.

The dashboard gives you a quick overview of the currently mapped topic. The statistics, in this case, are area-weighted. This can be changed on the Statistics tab. The first “fuel gauge” indicates how much of the map is used in the statistics. In this case it is 100%. If some polygons had no data for the topic then they would not be included in the statistics and the gauge would read the percentage area that had at least some of the needed data (something less than 100%).

The second gauge indicates how much of the needed data was available in the areas that were used for the statistics. In this case, again, it is 100%.

Right of the second gauge is a histogram showing the distribution of values in an area-weighted fashion. The horizontal line indicates the mean value. The circle represents the area-weighted mean for the map. In other words it is the average value, but only for those polygons included in the statistics. Remember that the polygons without data for the topic are NOT included in these statistics!

The legend area has two parts: a tabbed pane and a traditional legend. The traditional legend is the one we imported from MapWindow GIS on the last page of the setup. Each tab has a unique purpose.

Click on the Outline tab and navigate the outline to Cohesiveness (it's under Social Factors) and click on it. Now you should see a map with some green in it (figure 18). Note how the change in topic is also reflected in the Dashboard. The Outline tab lets you browse the knowledge base in a hierarchical fashion.

Click back to the Legend tab. Notice that now the previous topic button is enabled. The previous topic button and next topic button work like the corresponding buttons in your web browser: they let you move back and forth through your browsing history. The next set of buttons are menus for selecting dependency networks , calculated data links , and data links . The next two buttons let you jump to direct antecedents and direct dependents of the current topic, respectively. The final button in the toolbar is the options menu .

Now click in the western-most polygon (it's the community Rush). GeoNetWeaver should now look like figure 19. Note that now there is a cross-hair in the legend. Its up-down-ness indicates its value and its left-right-ness indicates its data sufficiency (how much of the data that is needed is actually there to be used).

Click the view button . This opens the dependency network for current topic, Cohesiveness (figure 20). Notice it is loaded with the data and results for the community we selected (Rush).

The results tab displays all of the results topics in a list (figure 21) along with their current values (at the moment for Rush). The currently mapped topic will be highlighted. Clicking on a topic in the list or clicking the map button will open that topic's map. Clicking the view button will open the topic's network viewer.

The inputs tab displays all of the inputs (data links) in a list (figure 22) along with their current values. Here the buttons work slightly differently. The map button will generate a map for the topic. Clicking the view button will initiate a search for all instances in the knowledge base where the data link is used.

The balance of the tabs are viewers for each of the data sets (Communities, Regions, Countries) (figures 23 - 25).