[Sigmetrics] Co-word Maps and Topic Modeling: A Comparison from a User's Perspective
dwojick at craigellachie.us
Fri Nov 13 10:22:25 EST 2015
Very interesting, Loet. Two thoughts:
If you want to visualize the argument in a document, that is what my issue
tree method does. See for example http://www.stemed.info/Repo_Tree.pdf.
However, this visualization is based on the logical relations between the
asserted propositions, which no algorithm that I know of can find. It would
be fun to try to develop one.
Second, the visualizations in your paper are two dimensional projections of
a complex three dimensional structure. Other projections (and other
structurings) might show other clustering, including possibly making the
topical modeling more meaningful.
At 01:14 AM 11/11/2015, Loet Leydesdorff wrote:
>PS. I forgot to provide the link at
><http://arxiv.org/abs/1511.03020>http://arxiv.org/abs/1511.03020; apologies. L.
><http://arxiv.org/abs/1511.03020>Co-word Maps and Topic Modeling: A
>Comparison from a User's Perspective
>Induced by "big data," "topic modeling" has become an attractive
>alternative to mapping co-words in terms of co-occurrences and co-absences
>using network techniques. We return to the word/document matrix using
>first a single text with a strong argument ("The Leiden Manifesto") and
>then upscale to a sample of moderate size (n = 687) to study the pros and
>cons of the two approaches in terms of the resulting possibilities for
>making semantic maps that can serve an argument. The results from co-word
>mapping (using two different routines) versus topic modeling are
>significantly uncorrelated. Whereas components in the co-word maps can
>easily be designated, the coloring of the nodes according to the results
>of the topic model provides maps that are difficult to interpret. In these
>samples, the topic models seem to reveal similarities other than semantic
>ones (e.g., linguistic ones). In other words, topic modeling does not
>replace co-word mapping.
>Loet Leydesdorff and Adina Nerghes
>** apologies for cross-postings
>Professor, University of Amsterdam
>Amsterdam School of Communication Research (ASCoR)
><mailto:loet at leydesdorff.net>loet at leydesdorff.net ;
>Honorary Professor, <http://www.sussex.ac.uk/spru/>SPRU, University of
>Guest Professor <http://www.zju.edu.cn/english/>Zhejiang Univ., Hangzhou;
>Visiting Professor, <http://www.istic.ac.cn/Eng/brief_en.html>ISTIC, Beijing;
>Visiting Professor, <http://www.bbk.ac.uk/>Birkbeck, University of London;
>SIGMETRICS mailing list
>SIGMETRICS at mail.asis.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the SIGMETRICS