Skip to contents

Computes pairwise cosine similarity between skills using their profession profile vectors. Skills appearing in fewer than min_professions are excluded before computation.

Usage

compute_skill_similarity(mat, min_professions = 3L)

Arguments

mat

Numeric matrix (rows = skills, columns = professions). Can be dense or sparse.

min_professions

Integer, minimum number of professions a skill must appear in (non-zero entries) to be included. Default: 3L.

Value

Dense symmetric matrix of skill-skill cosine similarities.