Compare Map User's Guide.
by Jana Diesner
CompareMap
Maps generated with AutoMap can be compared by using CompareMap.
1. Open CompareMap
In order to open CompareMap, click the Tools menu
in AutoMap and select Open CompareMap.
This is the CompareMap interface (CP notations are only used in this
picture for further reference to these panels throughout this help file):
2. Open a set of maps
All maps you wish to compare need to be stored in one folder. This folder
can also contain files in other formats.
Click the File menu and select Open multiple maps.
A file chooser will pop up. Choose the folder that contains the maps you wish
to analyze. This is typically the output folder in the root directory of AutoMap.
The folder that you select can contain files in multiple formats, CompareMap
automatically detects all .map files in this folder. Single click on this
folder. The folder will be highlighted in blue. Do not double click the folder.
Hit the
Open button. The names of the .map files contained in the selected folder will
be
displayed
on
CP1,
Map Selection index card. You can display the rest of the texts by using
the Browse Menu.
3. Select maps for comparison or consolidation
After you have opened a set of maps, choose the maps
you wish to compare at once by copying filenames from the left box to the right
box on CP1,
Map Selection index card by using the > button. You can move filenames
back
by
using
the
<
button. All files displayed in the right box will be considered for map comparison.
4. Comparison Settings
On CP2, in the Result type frame
on the Comparison
Settings index
card choose
either "count" or "binary" as the type of results by
clicking on the radio button. The default setting is "count".
5. Output Options
5.1 Computation of Hamming Distances
The
Hamming Distance is a measure for the distance of given pair of graphs from
each other. CompareMap offers two ways for computing Hamming Distance in dependence
of the type of map comparison that is run:
To compute the Hamming Distance for 1. and 2. check the Compute Hamming Distance box in the Hamming Distance field on CP2. The result will be printed out on CP4 on the Hamming Distance index card. A .csv with the Hamming Distance values is saved automatically in the output CompareMap folder under the root directory of AutoMap.
5.2 Output maps in DL format
Verbal Sets and consolidated maps are output in map format by default. In addition
to that, all maps can be converted into DL format automatically if requested
by the user. To generate DL files in addition to map files check the Output
in DL format check
box in the Output Options field on CP2.
The DL files are stored automatically in the DL_UciNet_Format folder
under the root directory of AutoMap.
6. Run Comparison
In order to run the map comparison, click the Run Comparison
menu and select on Run Comparison.
7. Consolidate maps
Purpose: Find the shared information of a set of maps.
The percentage of the shared information is determined by the user by choosing
a consolidation options on CP2, in
the Consolidation
of maps frame
on the Comparison Setting panel:
The union option equals the 0 percent of consolidation, which
equals the Union of Statements
If the lossy intersection option is chosen, a value between
5 and 95 percent needs to be chosen from the choose
percentage. As a result, the chosen percentage of shared knowledge
between the selected maps will be computed.
The intersection option equals
equals
the 100 percent of consolidation, which equals the Intersection of Statements.
8. Run Map Consolidation
In order to run the map consolidation,
click the Run Comparison menu
and select Consolidate Maps.
The resulting output file is stored automatically in the output CompareMap folder
under the root directory of AutoMap. To generate DL files check the Output
in DL format check box in the Output Options field
on CP2.
The DL files are stored automatically in the DL_UciNet_Format folder
under the root directory of AutoMap.
9. Outputs
The outputs of a Map Comparison are
displayed on CP3 on the Verbal Sets index
cards named Union of Statements, Consensus of Statements, and Dissension
of Statements and on
CP4, Stat, Statements index
card.
The map files are stored automatically in the output CompareMap folder under
the root directory of AutoMap.
The outputs of a Map Consolidation are
displayed on CP3 on the Verbal Sets index
card named Consolidated Maps. No stat output for map consolidation
is computed.
The map file is stored automatically in the output CompareMap folder
under the root directory of AutoMap.
Outputs of both types of analysis are automatically saved in folder
named “output CompareMap” under directory where AutoMap 2.0 is
installed. Two types of files are generated: map files and stat files:
9.1 Map
Three map files are displayed on CP3:
Each map contains one statement per line.
If count was chosen as the comparison setting the first column of the Union
map and Consensus map indicates the frequency of
every displayed statement. The Dissension map is binary by default.
If binary was chosen as the comparison setting the maps do not indicate
the frequency of displayed statement. The Dissension
map is binary by default.
Each map generated is automatically saved as Union.map, Consensus.map, and
Dissension.map in folder named “output CompareMap” under
directory where AutoMap 1.2 is installed.
9.2 Statistics
After running the analysis, the Stat file will be displayed on CP4
,Stat, Statements Index Card.
Each stat file generated is automatically saved as CompareStatements.stat in
folder named “output
CompareMap” under directory where AutoMap 1.2 is installed.
10. Example for Comparison Outputs
Input texts:
Denmark.txt: Reporters said hundreds of people emerged from
shops in Copenhagen city centre to see what was happening, and used their mobile
phones to contact their families.
USA.txt: The NYPD said a number of
people were trapped in elevators for awhile. Thousands of people left buildings
and
walked into the streets.
Delete List: of, from, in, to, what, was, and, their, a, were, for, awhile,
into
Thesaurus: reporters/ agents;
nypd/agents;
hundreds/number;
emerged/ left;
shops/buildings
Resulting maps for Denmark.txt and USA.txt that were used for
the comparison:
| denmark.txt.map: | usa.txt.map |
 |
 |
Map comparison was run with the count setting on both maps.
Resulting maps:
| Union of Statements: | Consensus of Statements: | Dissension of Statements: |
 |
 |
 |
Resulting stat file:
# of statements in union:
unique: 20
total: 25
# of statements in consensus:
unique: 5
total: 5
# of statements in dissension:
unique: 15
Groups identified by CompareMap aim to represent clusters of similar maps. The clusters suggested by CompareMap are exclusive but not exhaustive. This means that each map belongs to either one or none cluster. As a baseline for determining similarity between texts CompareMap uses Hamming Distances. The user needs to provide CompareMap with a threshold value of Hamming Distance below that maps are considered for cluster identification. To pick that threshold:
12. Exit
To exit CompareMap, click the File menu and
select Exit.
CompareMap will be closed.
Since CompareMap is a part of the AutoMap package, this reference can be used:
Carley, K.M., & Diesner, J. (2005). AutoMap: Software for Network Text
Analysis. CASOS (Center for Computational Analysis of Social and Organizational
Systems), ISRI, CMU.
For further information on CompareMap please visit: http://www.casos.cs.cmu.edu/projects/automap
On this web page you will find:
We also provide a online
discussion forum for AutoMap, which also covers CompareMap. The purpose
of this forum is to:
- Discuss questions related to the AutoMap software.
- Get help from other AutoMap users and the developers of the software with
using the tool.
Questions, Bugs, and Comments
Please contact us:
Dr. Kathleen M. Carley (kathleen.carley@cmu.edu)
Jana Diesner (diesner@cs.cmu.edu)
Carnegie Mellon University
School of Computer Science
Institute for Software Research International (ISRI)
Center for Computational Analysis of Social and Organizational Systems (CASOS)
5000 Forbes Avenue
1325 Wean Hall
Pittsburgh, PA, 15213