Sci2 Manual : 5.2.1 Funding Profiles of Three Universities (NSF Data)
This page last changed on Mar 16, 2011 by dapolley.
Load 'Cornell.nsf', 'Michigan.nsf', and 'Indiana.nsf' using 'File > Load" and following this path: 'yoursci2directory/sampledata/scientometrics/nsf'. Use the following workflow for each of the three nsf files loaded. Select each of the datasets in the Data Manager window and run 'Data Preparation > Extract Co-Occurrence Network' using the following parameters (Note that the Aggregation Function File is 'yoursci2directory/sampledata/scientometrics/properties/nsfCoPI.properties'): Two derived files per dataset will appear in the Data Manager window: the co-PI network and a merge table. In the network, nodes represent investigators and edges denote their co-PI relationships. (To learn how the merge table can be used to further clean PI names, see section 5.1.4.2 Author Co-Occurrence (Bibliographic Coupling) Network.) Choose the "Extracted Network on Column All Investigators" and run 'Analysis > Networks > Network Analysis Toolkit (NAT)' for each dataset. This will display the amount of nodes and edges, as well as the amount of isolate nodes that can be removed by running 'Preprocessing > Networks > Delete Isolates'. To see a more detailed view of any of the components in the network (e.g. the largest Indiana component) select the network with deleted isolates in the Data Manager: Then, run 'Analysis > Networks > Unweighted & Undirected > Weak Component Clustering' with the parameter: Indiana's largest component has 19 nodes, Cornell's has 67 nodes, and Michigan's has 55 nodes. Visualize Indiana's network in GUESS using the 'yoursci2directory/scripts/GUESS/co-PI-nw.py' script. Save the file as a jpg by selecting 'File > Export Image'. Use the "Browse…" option in the "Export Image – GUESS" popup window to select the folder in which you would like to save the image. Figure 5.24: Largest component of Indiana University co-PI network. Node size and color display the total award amount. 5.2.1.1 Database Extractions
The Sci2 Tool supports the creation of databases for NSF files. Database loading improves the speed and functionality of data preparation and preprocessing. To load the Indiana NSF file as a database, go to 'File > Load > and select yoursci2directory/sampledata/scientometrics/nsf/Indiana.nsf.' In the "Load" pop-up window, choose "NSF database." Cleaning should be performed before any other task using 'Data Preparation > Database > NSF > Merge Identical NSF People'. To create Co-PI networks like those from the previous workflows, simply run 'Data Preparation > Databases > NSF > Extract Co-PI Network' on the cleaned database. Delete the isolates by running 'Preprocessing > Networks > Delete Isolates'. Figure 5.25: Indiana University Co-PI network using databases. ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
![]() |
Document generated by Confluence on May 31, 2011 15:16 |