Build a master skills list with metadata and diffusion classification
Source:R/relevance.R
build_skillist.RdAggregates skill metadata (reuse type, flag columns) and merges with diffusion scores. The ESCO reuse type labels are translated to Italian.
Arguments
- skills
A
data.tableof skill occurrences with required columnsescoskill_level_3,esco_v0101_reusetype,esco_v0101_green,esco_v0101_language, and the optional columnspillar_softskillsandesco_v0101_ict.- diffusion
A
data.tableas returned bycompute_skill_diffusion(), with columnsescoskill_level_3,N,tf,idf,diffusione.
Value
A data.table with one row per unique skill, including metadata
columns, the Italian type label (tipo), and diffusion scores. The
columns pillar_softskills and esco_v0101_ict are always present, and
are NA_integer_ when absent from skills.
Details
The tipo classification derives from esco_v0101_reusetype
alone; pillar_softskills and esco_v0101_ict are descriptive
pass-through flags. They are optional because the Lightcast data_v2
delivery consumed through itaposts no longer supplies them: data read
with read_oja_itaposts() yields NA in both columns, while
legacy sources that still carry the flags pass them through unchanged.