divergence

Compute the divergence of topics. This can be used to search the optimal
number of topics for LDA.

Implements the seeded-LDA model (Lu, Ott, Cardie & Tsou 2010) <doi:10.1109/ICDMW.2011.125> using the quanteda package and the GibbsLDA++ library for semisupervised topic modeling.
Seeded-LDA allows users to pre-define topics with keywords to perform theory-driven analysis of textual data in social sciences and humanities (Watanabe & Zhou 2020) <doi:10.1177/0894439320907027>.

Kohei Watanabe

seededlda

Seeded-LDA for Topic Modeling

Phan Xuan-Hieu

divergence function

<dl><dt>x</dt>
<dd>a LDA model fitted by <code>textmodel_seededlda()</code> or <code>textmodel_lda()</code>.</dd>
<dt>weighted</dt>
<dd>if <code>TRUE</code> weight the divergence scores by the sizes of
topics.</dd>
<dt>min_size</dt>
<dd>the minimum size of topics that can increase the average
divergence. Ignored when <code>weighted = FALSE</code>.</dd>
<dt>select</dt>
<dd>names of topics for which the divergence is computed.</dd></dl>

Arguments

[Experimental] Compute the divergence of topics — divergence

<dl>

<dt>x</dt>
<dd>a LDA model fitted by <code>textmodel_seededlda()</code> or <code>textmodel_lda()</code>.</dd>


<dt>weighted</dt>
<dd>if <code>TRUE</code> weight the divergence scores by the sizes of
topics.</dd>


<dt>min_size</dt>
<dd>the minimum size of topics that can increase the average
divergence. Ignored when <code>weighted = FALSE</code>.</dd>


<dt>select</dt>
<dd>names of topics for which the divergence is computed.</dd>

</dl>

divergence: [Experimental] Compute the divergence of topics

Description

Usage

Arguments

Details

References

See Also