Topics have no fixed order: topic 3 of one run may be topic
12 of another. The topics of model_b are matched one-to-one to those of
model_a so as to maximise the total cosine similarity of their word
distributions (assignment problem, Hungarian method; Kuhn 1955), as in
the stability analyses of Greene et al. (2014) and Roberts, Stewart and
Tingley (2016).
Usage
match_topics(model_a, model_b, n_terms = 10L, ...)
Arguments
- model_a, model_b
Fitted models with the same number of topics and the
same vocabulary.
- n_terms
Number of top words compared for the jaccard overlap.
- ...
Not used: an argument renamed in the development version of mintR gives an error
naming its replacement, any other one an error.
Value
A data.table with topic_a, topic_b, cosine (similarity of
the word distributions) and jaccard (overlap of the n_terms top
words: shared / union).
References
Kuhn, H. W. (1955). The Hungarian method for the assignment problem.
Naval Research Logistics Quarterly, 2(1-2), 83-97.
