Integrating skill sets: mapping industry expectations of graduate mechanical engineers in Australia via network analysis of job advertisements

Graduate employability is a critical issue in Australian engineering education, where high competition for entry-level roles persists despite industry skill shortages. While Engineers Australia’s Stage 1 Competency Standard provides a benchmark, there is limited empirical evidence defining the specific skill sets required for graduate mechanical engineers. This study addresses this gap by analysing a national dataset of job advertisements to identify and map priority skills. Unlike automated studies producing broad skill inventories, this research delivers high-resolution, discipline-specific mapping of skill interdependencies. By analysing a manually verified dataset through Louvain community detection and Apriori association rule mining, the research reveals the structural architecture of employability in mechanical engineering. Findings indicate that employers seek integrated skill sets rather than isolated technical abilities. Proficiency in CAD and computational software is fundamental but consistently demanded alongside professional skills like communication, teamwork, and project management. Clustering analysis identified distinct role archetypes combining technical, professional, and contextual attributes. Rule mining uncovered strong associations between software expertise, stakeholder communication, and understanding ethical and sustainable practices. This research demonstrates that employability is constructed through holistic integration. Consequently, engineering curricula must evolve to embed these integrated skills, ensuring educational outcomes align directly with contemporary industry expectations.

毕业生就业能力是澳大利亚工程教育领域的关键议题。尽管行业存在技能短缺问题，但入门级岗位的竞争依然激烈。澳大利亚工程师协会（Engineers Australia）的第一阶段能力标准虽提供了基准参考，但目前针对机械工程应届毕业生所需的具体技能组合，尚缺乏充足的实证依据。本研究通过分析全国范围的招聘广告数据集，识别并梳理优先技能，以此填补这一研究空白。与生成宽泛技能清单的自动化研究不同，本研究实现了精细化、学科专属的技能相互依赖关系图谱绘制。本研究借助卢万社区检测（Louvain community detection）与Apriori关联规则挖掘（Apriori association rule mining）算法，对经过人工验证的数据集展开分析，揭示了机械工程领域就业能力的结构体系。研究结果显示，雇主青睐整合式的技能组合，而非孤立的单项技术能力。熟练掌握计算机辅助设计（CAD）与计算类软件是基础要求，但雇主始终同时要求求职者具备沟通、团队协作与项目管理等专业技能。聚类分析识别出了融合技术、专业与情境属性的差异化岗位原型。关联规则挖掘则揭示了软件专业能力、利益相关方沟通与道德及可持续实践认知之间的强关联。本研究表明，就业能力是通过全面整合的能力体系构建而成的。因此，工程类课程体系必须进行革新，融入这些整合式技能，确保教育成果与当代行业预期直接匹配。