The general Sci2 Tool user interface is shown in Figure 2.3.
Figure 2.3: Sci2 Tool Interface Components
The Sci2 Tool menu structure is arranged such that a workflow runs from left to right. The 'File' menu on the left allows a user to load data in a number of formats, which can then be prepared, preprocessed, analyzed, and finally visualized. Users also have the option of modeling new networks or finding help online.
In the 'File' menu, the first option after clicking on 'Load' is 'Select a File':
Figure 2.4: Using 'Load' to select a file in Sci2
The 'File' menu functionality includes loading multiple data formats (see section 2.3 Data Formats for details), loading ISI and NSF data into a database, saving and viewing results, and merging or splitting node and edge files. 'Load and Clean ISI File' automatically normalizes author names and merges duplicate records, and is specifically designed for text-based scientometric workflows (algorithms within 'Data Preparation > Text Files'). For database manipulation of ISI or NSF files, use 'File > Load ...' A pop-up window will appear allowing you to select what sort of file you are importing (Scopus or ISI csv, for example) or how you want the Sci2 Tool to read the file ('ISI scholarly format' or 'ISI database,' for example).
Figure 2.5: The 'Load' pop-up window
The 'Console' window documents all operations performed on the data.
Figure 2.6: Sci2's Console and Schedule after successfully loading a database file
When the scheduler indicates that the 'Load' operation is complete, the .isi file will appear in the data manager, preceded by a database icon. The converter graph and directory reader produces a sample graph based on filetypes supported by the Sci2 Tool and a sample tree based on any directory structure on the hard drive, respectively.
After loading a file, use options in the 'Data Preparation' menu to clean the data and create networks or tables which can be used in the preprocessing, analysis, and visualization steps. The 'Data Preparation > Database' menu is specifically for ISI or NSF data previously loaded into a database. Options in 'Data Preparation > Text Files' are for any table-based datasets (like csv files) and are used to extract networks. Find detailed information on each menu item in section 3.1 Sci2 Tool Plugins.
Figure 2.7: Data Preparation options
Use preprocessing algorithms to prune or append networks or tables before analyzing and visualizing them. The menu is separated by domain, and most simple tasks require staying within the same domain. For example, to visualize a co-authorship network, only use algorithms within the 'Networks' domain under 'Preprocessing', 'Analysis', and 'Visualization'. Similarly, a geographic map requires only 'Geospatial' algorithms. Find detailed information on each menu item in section 3.1 Sci2 Tool Plugins.
Figure 2.8: Preprocessing options
Once data is loaded, prepared, and processed with whatever features needed, analysis is possible in each of the four domains: temporal, geospatial, topical, or network.
Figure 2.9: Analysis options
Analysis results can be used on their own or in conjunction with visualizations to gain insight into a dataset. The Sci2 Tool features predominantly network analysis algorithms, however the tool also supports geocoding of table data and burst analysis for topical or temporal studies. Find detailed information on each menu item in section 3.1 Sci2 Tool Plugins.
The Sci2 Tool supports the creation of new networks via pre-defined models. Learn more about modeling in section 4.10 Modeling (Why?).
Figure 2.10: Modeling options
Once all previous data steps are complete, the Sci2 Tool can visualize the results. The most popular choice for visualizing networks is the GUESS toolkit, or DrL for much larger scale networks. Geocoded data can be represented on a map of the world or the United States, and temporal or topical data can be viewed using the horizontal bar graph. Find detailed information on each menu item in section 3.1 Sci2 Tool Plugins.
Figure 2.11: Visualization options
The 'Help' menu leads to online documentation, advanced tool configuration, and detailed development information.
Figure 2.12: Help options
All operations such as loading, viewing, or saving datasets, running various algorithms, and algorithm parameters, etc. are logged sequentially in the 'Console' window as well as in log files stored in the 'yoursci2directory /logs' directory. The Console window also displays the acknowledgement information about the original authors of the algorithm, the developers, the integrators, a reference paper, and the URL to the reference if available, together with an URL to the algorithm description in the NWB/Sci2 community wiki.
The 'Data Manager' window displays all currently loaded and available datasets. The type of a loaded file is indicated by its icon:
Text – text file
Table – tabular data (csv file)
Matrix-data (Pajek .mat)
Plot – plain text file that can be plotted using Gnuplot
Network – Network data (in-memory graph/network object or network files saved as Graph/ML, XGMML, NWB, Pajek .net or Edge list format)
Database – In-memory database
Tree – Tree data (TreeML)
Derived datasets are indented under their parent datasets. That is, the children datasets are the results of applying certain algorithms to the parent dataset.
The 'Scheduler' lets users keep track of the progress of running algorithms.