Gold-Standard Labelled Job Postings for Information Extraction Training

Jobseek Postings Labelled is a collection of gold-standard labelled job postings sampled daily from public company career pages. It is produced by a Claude-Code-orchestrated pipeline with specialized Sonnet subagents for section splitting and per-section extraction. The dataset serves as the substrate for training an improved structured-information extractor for jseek.co.

Use Cases

• Train a job information extractor based on the gold-standard labels.
• Benchmark NLP models for section splitting and named entity recognition based on the specialized subagent annotations.
• Analyze trends in job postings based on the daily sampling methodology.
• Develop models for normalizing structured fields from unstructured job descriptions based on the per-section extraction labels.

Strengths

• Gold-standard labelled data, implying high annotation quality.
• Sampled daily, suggesting a continuous update process.
• Produced by a specialized AI pipeline with Claude-Code and Sonnet subagents.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Public company career pages.

Collection Method

Sampled daily via a Claude-Code-orchestrated pipeline with specialized Sonnet subagents.

Freshness

Last updated 2026-05-09 18:21:22; freshness should be verified.