Skills required in ADEM job vacancies

This dataset contains data on the skill requirements identified in each job vacancy that was transferred to ADEM’s database between 2015 and now. Employers in Luxembourg are legally obliged to transfer their open job vacancies to ADEM. Each row of the dataset corresponds to a vacancy-skill combination: for example, if vacancy_id=”123” and skill=”work in teams”, this means that our algorithm identified the skill of working in teams in the description of vacancy “123”. If the column skill is empty for a given vacancy, this means that no skill requirement could be identified for that vacancy. Skills are identified through an algorithm that automatically reads the job vacancy description and classifies the skill requirements identified based on fuzzy text matching, given a set of keywords corresponding to each skill potentially required. Identified skills are structured according to the ESCO framework, based on the metadata made available (see https://esco.ec.europa.eu/en/classification/skill_main). Most other variables (number of advertised job positions, canton of work location, occupational code) are based on information declared by the employers transferring the vacancy to ADEM; the month and year refer to the time of publication. The algorithm for skill identification is currently still under development, and a more detailed methodological guide will be published when the methodology will become more stable. Variables Vacancy_id: Dataset-specific Id of the vacancy Skill: ESCO preferred label for the skill (see https://esco.ec.europa.eu/en/classification/skill_main) Skill_uri: ESCO id for the skill (see https://esco.ec.europa.eu/en/classification/skill_main) Positions: open positions advertised in the vacancy (e.g. if a company opens a vacancy to recruit 7 accountants, then positions = 7) Month: month of publication of the vacancy Year: year of publication of the vacancy Canton: Canton in which the advertised job would be located Occupation_code: ROME occupational code that best fits the vacancy (see https://rome.adem.etat.lu/index_base.html and https://www.francetravail.fr/employeur/vos-recrutements/le-rome-et-les-fiches-metiers.html) Occupation_label: ROME occupational code label for the vacancy (see https://rome.adem.etat.lu/index_base.html and https://www.francetravail.fr/employeur/vos-recrutements/le-rome-et-les-fiches-metiers.html)

本数据集收录了自2015年起至现今，ADEM数据库中各职位空缺所标识的技能要求数据。卢森堡的雇主依法须将其开放的职位空缺信息提交至ADEM。数据集的每一行对应一个职位技能组合：例如，若职位ID为“123”，技能为“团队合作”，则表示我们的算法在职位“123”的描述中识别出了团队合作这一技能。若某职位行中的技能列空白，则意味着该职位空缺无法识别出技能要求。技能通过算法自动读取职位空缺描述，并根据对应技能可能需求的词汇集进行模糊文本匹配进行分类识别。识别出的技能按照ESC框架进行结构化，基于提供的元数据（详见https://esco.ec.europa.eu/en/classification/skill_main）。大多数其他变量（如广告职位数量、工作地点的州、职业代码）基于提交至ADEM的雇主所提供的信息；月份和年份指职位发布的时期。技能识别算法目前仍在开发中，当方法论趋于稳定时，将发布更详细的指导手册。变量包括：Vacancy_id：数据集特定的职位ID；Skill：技能的ESC推荐标签（详见https://esco.ec.europa.eu/en/classification/skill_main）；Skill_uri：技能的ESC ID（详见https://esco.ec.europa.eu/en/classification/skill_main）；Positions：职位空缺中广告的开放职位数量（例如，若一家公司为招聘7名会计师而开放职位，则positions = 7）；Month：职位空缺发布的月份；Year：职位空缺发布的年份；Canton：广告职位所在州的名称；Occupation_code：最适合职位的ROME职业代码（详见https://rome.adem.etat.lu/index_base.html和https://www.francetravail.fr/employeur/vos-recrutements/le-rome-et-les-fiches-metiers.html）；Occupation_label：职位的ROME职业代码标签（详见https://rome.adem.etat.lu/index_base.html和https://www.francetravail.fr/employeur/vos-recrutements/le-rome-et-les-fiches-metiers.html）。