By applying a statistical technique called principal-component analysis to the distribution of topics, they were able to divide modern, popular music into 13 genre-straddling "styles" (one, for example, draws from "rock", "classic rock", "pop" and "new wave" while another draws from "rock", "hard rock", "alternative" and "classic rock").
It shows preliminary results about the distribution of topics in Automatic Control teaching.
The LDA model assumes that a document collection is generated by distribution of topics, and each topic is a probabilistic distribution of terms.
Figure 6 combines the topic composition of all analysed documents and shows the distribution of topics for the combined document corpora.
For our analysis we were furthermore interested in the distribution of topics between the "citizen science" and "forest monitoring" document corpora.
The distribution of topics to new document is given by the following expression P ( z k | d n e w ) = ∑ w i ∈ W n ( d n e w, w j ) P ( z k | d n e w, w j ) n ( d n e w ).
In PLSA, this is achieved by keeping the distribution of words that make up a topic (P(w j |z k )) fixed while the distribution of topics that new document belong to (P z k |dnew)) is adapted at each M-step.
As DBLP does not have influence probabilities from the real data, we simulate two topics according to the joint distribution of topics 1 and 2 in the Flixster and follow the practice of the TRIVALENCY model in [10] to rescale it into 0.1, 0.01, or 0.001, standing for strong, medium, and low influence, respectively.
The frequency distribution of topics taught was not significantly different between institution types (comparison not shown, 3-way Chi Square comparison made) [not addressed X 2 (57, N = 438) = 49.1; addressed superficially X 2 (57, N = 438) = 40.52; addressed significantly X 2 (57, N = 438) = 21.66, p > 0.05].
