Mapping AI Job Skill Bundles: Frequency, Co-Occurrence, and Market Basket Analysis of 500 AI Job Postings

This study analyzes 500 AI-related job postings collected from PostJobFree (2025) using a curated taxonomy of 120 employer-facing skill terms, with 88 technical skills spanning AI/ML methods, programming languages, data tools, analytics, data-engineering platforms, and cloud/MLOps infrastructure, plus 32 non-technical skills covering communication, leadership, and organizational competencies. A hybrid keyword–regular expression pipeline produces two complementary frequency measures: a mention-level count (11,250 total mentions) and a posting-level count (4,106 occurrences). All 120 skills are reported in frequency analyses; co-occurrence and association-rule analyses are restricted to the 88-term technical subset to preserve analytical signal. Results show that core AI and machine-learning skills, cloud platforms, and data-engineering tools recur in stable bundles, and that strong association rules link cloud platforms, deep-learning frameworks, and SQL. Grounded in human capital theory and the principle of skill complementarity, the findings demonstrate that AI hiring is structured around coherent skill bundles rather than isolated requirements, providing empirical grounding for AI curriculum design, workforce planning, and program learning outcomes.