Anne Kennedy wrote:
>
> Does anyone have any experience with TextSmart (a relatively new SPSS
> product for handling textual data)?
we tested TextSmart Version 1.1 and we would suggest that the
automatic generation of categories can be problematic:
The program builds categories by means of cluster analysis
based on single words and lemmas (grouped as alias).
The quality of these automatically built categories depends on
on how good the stop word list and the own alias lists are.
We have found out that one has to invest a lot of time to build
'meaningful' stop word and alias lists. Furthermore, even
if one has invested a lot of effort in doing this, it is not
easy to interpret the meaning of the generated categories.
However, one can switch off this functionality and build one's own
categories manually (always based on words or aliases).
We found the manual very good, and as far as the implementation is
concerned, version 1.1 is more stable and faster than version 1.0, but
still
not fast enough.
If you are interested, our review should be made public some time next
month.
Hope this helps.
Best
Cornelia Zuell and Melina Alexa
> --
> Dr. Harald Klein
> TU Ilmenau
> Media and communication research
> PF 100 565
> 98684 Ilmenau
> Germany
> Tel.: +49 3677 694672
> Tel/Fax: +49 36783 80284
--
-----------------------------------------------------------------
Cornelia Zuell
ZUMA
P.O. Box 12 21 55
D-68072 Mannheim
Germany
Tel: +49 6211246147, Fax: +49 6211246100
e-mail: [log in to unmask], Internet: http://www.zuma-mannheim.de
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|