TY - COMP
T1 - Clusterizor
T2 - An R-based cluster analysis program for linguistic analysis
A2 - Jensen, Kim Ebensgaard
N1 - V 1.1b
Disclaimer:
Program and its output come with ABSOLUTELY NO WARRANTY, and the user is fully responsible for all aspects; all consequences and costs resulting from use, or inability to use, are assumed by the user. In no event are copyright holders, developers, modifiers or distributors liable to the user nor responsible for any consequences resulting from use or inability to use Clusterizor.
Distribution:
Clusterizor is free software and may redistributed and/or modified under the terms of the GNU General Public License as published by the Free Software Foundation version 2 or later. Note that Clusterizor is intended solely for scientific and research-related use and should under no circumstances be used to serve commercial purposes.
Running Clusterizor:
Type, or copy-paste, the following into R:
source("http://vbn.aau.dk/files/77738073/Clusterizor.r")
clusterizor()
Then follow the instructions.
Use the attached Test Input File for experimentation. Download it to your computer and use it as input file (it's called 'test.txt').
Projected updates:
Clusterizor is work in progress and, while operational, not fully developed yet. Implementation of the following improvements is planned:
- more dendrogram manipulation
- streamlining/improvement of user-program interaction
- general elegance
PY - 2013/4/9
Y1 - 2013/4/9
N2 - Clusterizor is a cluster analysis program for linguistic analysis. Clusterizor is, in its current state, restricted to hierarchical cluster analysis, with other cluster analysis types currently being on the drawing board. Clusterizor allows you to generate dendrograms on the basis of combinations of binary, Canberra, Euclidean, Manhattan City Block, maximum, and Minkowski distancings with average, centroid, furthest-neighbor, nearest-neighbor, McQuitty, median, and Ward clustering methods. With those distancing types that, involve decimals in distance matrices, you may choose between one and five decimals. It also allows you to export distance matrices in tabular form. You may also choose to include a grid and red boxes in the dendrogram.Clusterizor first runs the user's chosen combination of distancing and clustering methods. Then a distance matrix is generated on the basis of the user's input file, in accordance with the chosen distancing method. The user will be given the option to export the distance matrix as a text file. The chosen clustering method is applied to the distance matrix, resulting in a dendrogram illustrating the cluster relations.
AB - Clusterizor is a cluster analysis program for linguistic analysis. Clusterizor is, in its current state, restricted to hierarchical cluster analysis, with other cluster analysis types currently being on the drawing board. Clusterizor allows you to generate dendrograms on the basis of combinations of binary, Canberra, Euclidean, Manhattan City Block, maximum, and Minkowski distancings with average, centroid, furthest-neighbor, nearest-neighbor, McQuitty, median, and Ward clustering methods. With those distancing types that, involve decimals in distance matrices, you may choose between one and five decimals. It also allows you to export distance matrices in tabular form. You may also choose to include a grid and red boxes in the dendrogram.Clusterizor first runs the user's chosen combination of distancing and clustering methods. Then a distance matrix is generated on the basis of the user's input file, in accordance with the chosen distancing method. The user will be given the option to export the distance matrix as a text file. The chosen clustering method is applied to the distance matrix, resulting in a dendrogram illustrating the cluster relations.
KW - Cluster Analysis
KW - Linguistic Cluster Analysis
KW - R
KW - Quantitative Linguistic Analysis
KW - Software
KW - Dendrogram
KW - corpus linguistics
M3 - Computer programme
ER -