The classifier retains its trained network, vocabulary and topic keywords. Reading a table or predicting new documents never trains the network. Attention tables describe internal aggregation weights. They are not class-specific contributions or explanations of a prediction.
Usage
# S3 method for class 'hg_hypergat'
hg_get(
x,
what = NULL,
split = NULL,
correct = NULL,
node = NULL,
sort_by = NULL,
top = NULL,
...
)
# S3 method for class 'hg_hypergat'
print(x, ...)
# S3 method for class 'hg_hypergat'
summary(object, ...)
# S3 method for class 'hg_hypergat'
plot(
x,
y,
type = c("confusion", "history", "hyperedges", "attention"),
node = NULL,
top = 8L,
...
)Arguments
- x
A fitted
hg_hypergat.- what
NULLor"predictions","accuracy","classes","confusion","history","documents","attention","hyperedges"or"hyperedge_words". Evaluation tables use held-out documents only."documents"includes the text with its prediction.- split, correct, sort_by
Filters and ordering for predictions and documents, as in
hg_get.hg_classification().- node
Document ids or a table with a
nodecolumn.- top
Number of rows per class (predictions and documents), per document (hyperedges or attention), or per hyperedge (hyperedge words).
- ...
Unused.
- object
A fitted
hg_hypergat.- y
Unused.
- type
"confusion"(default),"history","hyperedges"(the word hypergraph of one document), or"attention". Attention plots show weights alongside their uniform normalization baseline. These weights do not measure a sentence's contribution to the predicted class.
Value
A plain data frame. The attention tables include uniform
normalization baselines: word_uniform = 1 / n_words within an edge,
and edge_uniform = 1 / n_edges for a word. Hyperedge summaries include
uniform_weight (mean edge_uniform) and exclusive_fraction (the
fraction of words belonging only to that hyperedge). Such words have
edge_weight = 1 regardless of the learned parameters.