|
|
Lebart Ludovic / DtmVic Directeur de recherches C.N.R.S. (R) |
|
|
|
|
Lebart Ludovic / DtmVic Directeur de recherches C.N.R.S. (R) |
|
|
| Page from the website: http://www.dtm-vic.com |
Software DtmVic: Exploratory statistical processing of complex data sets comprising both numerical and textual data.
Applications concern primarily the field of "Multivariate Descriptive Statistical Analysis", the processing of
responses to open ended questions in socio-economic sample
surveys, the exploration of texts (with or without metadata).
- Special emphasis on:
Complementary use of visualization techniques
(Principal Component Analysis, Two-way and Multiple
Correspondence Analysis) and clustering techniques (hybrid
method using both hierarchical clustering and k-means
technique; Self Organizing Maps (SOM); Additive trees.
Assessments of visualization techniques:
resampling techniques (bootstrap, partial bootstrap, total
bootstrap, bootstrapping variables). Three kinds of total
bootstrap are provided: Total bootstrap type 1: simple
change of sign of the axes for the replicated analyses,
when needed. Total bootstrap type 2: as type 1, plus:
corrections for possible interversions of axes. Total
bootstrap type 3: corrections of the replicated principal
subspaces using procrustean analyses.
Contiguity analysis and related methods.
Kohonen maps (SOM).
The present version of this free
software contains a series of typical examples of data
processing and textual data processing.
[13 commented application examples and 3 examples of
data importation are available].
The user can run the software on his/her own data sets via
changing a few parameters and complying with the input formats
shown in the examples (data, dictionary and texts).
Several importation procedures can make much easier the input
of data.
The menu item: "Create a command file" allows the user to
generate easily a command file.
Importation from Excel(r), exportation of dtm files to Excel or
R.
Limitations about this version: 30,000 respondents (individuals, rows), 1,200 variables (numerical or categorical), 100,000 characters for the response of an individual to a set of open-ended questions; (No theoretical limitation of size for the texts in the case of texts not inserted in a sample survey).
Last update of the software Dtm-Vic : February 6th, 2020)
