Taxonomies

Overview

Taxonomies are standardized classification systems that organize labor market concepts into consistent, reusable identifiers.

Lightcast MCP uses taxonomies to represent dimensions such as occupations, skills, companies, industries, education programs, schools, benefits, and geographic locations.

By using standardized identifiers instead of free-form text, taxonomies help ensure that filtering, grouping, ranking, and analysis produce accurate and consistent results.

When working with a taxonomy-backed dimension, use the classify tool to convert user-provided text into a canonical ID before using it in an analysis request.

Benefits of Taxonomies

Taxonomies provide a common language across Lightcast datasets and tools.

Benefit	Description
Normalization	Maps varied the user language to canonical IDs.
Consistency	Ensures the same concept is represented consistently across requests.
Validation	Helps ensure that requests use supported values and levels.
Aggregation	Enables filtering, grouping, and ranking using standardized concepts.

Classifiable Dimensions

The following dimensions are backed by taxonomies and support classification through the classify tool.

Dimension	Datasets	Levels
`benefit`	`job_postings`	1 → Benefit Category 2 → Benefit
`company`	`job_postings`, `profiles`	1 → Company
`education_program_cip`	`job_postings`, `profiles`	1 → 2-digit series 2 → 4-digit subseries 3 → 6-digit program
`education_school`	`profiles`	1 → School
`industry_naics`	`job_postings`, `profiles`, `core_lmi`	1 → Sector 2 → Subsector 3 → Industry group 4 → NAICS industry 5 → National industry
`location_area_us`	`job_postings`, `profiles`, `core_lmi`	1 → State 2 → MSA 3 → County
`location_us_city`	`job_postings`, `profiles`	1 → City
`occupation_lot`	`job_postings`, `profiles`	1 → Career area 2 → Occupation group 3 → Occupation 4 → Specialized occupation
`occupation_onet`	`job_postings`, `profiles`	1 → O*NET occupation
`occupation_soc`	`job_postings`, `profiles`, `core_lmi`	1 → Major group 2 → Minor group 3 → Broad occupation 4 → Detailed occupation
`postings_education_level`	`job_postings`	1 → Education level
`profiles_education_level`	`profiles`	1 → Education level
`skill`	`job_postings`, `profiles`	1 → Skill Category 2 → Skill Sub-Category 3 → Skill

Classification Workflow

Use classify whenever a user provides text for a taxonomy-backed dimension.

For example:

{
  "dimension": "occupation_soc",
  "level": 4,
  "value": "software developers"
}

The classify tool returns matching concepts along with their canonical identifiers, names, hierarchy levels, and optional confidence scores.

Use the selected identifier when constructing your filter:

{
  "occupation_soc": {
    "level": 4,
    "values": ["231715","2317"]
  }
}

Using taxonomy identifiers instead of raw user text helps ensure requests pass validation and return consistent results.

Taxonomy Hierarchies

Many taxonomies are hierarchical. The selected level determines how broad or specific a classification will be.

Taxonomy	Broader Level	More Specific Level
SOC	Major Group	Detailed Occupation
NAICS	Sector	National Industry
Skill	Skill Category	Skill
CIP	2-Digit Series	6-Digit Program
US Area	State	County

Choose the level that best matches the question being asked.

Higher-level classifications support broader analysis and aggregation, while lower-level classifications provide more detailed and targeted results.