lisc.analysis.counts.compute_association_index¶
- lisc.analysis.counts.compute_association_index(data, counts_a, counts_b)[source]¶
Compute the association index from the co-occurrence data.
- Parameters:
- data2d array
Counts of co-occurrence of terms.
- counts_a, counts_b1d array
Counts for each individual search term.
- Returns:
- index2d array
The association score of the co-occurrence data.
Notes
This computes the Jaccard index, as \(AI_{ij} = |c_{ij} N d_{ij}| / |c_{ij} U d_{ij}|\).
The denominator, \(|c_{ij} U d_{ij}|\), is equivalent to \(|c_{ij}| + |d_{ij}| - |c_{ij} N d_{ij}|\).
