Progetti di ricerca

AI4ESCO aims at realising the objective (a) to support Member States, EURES members and partners to establish high-quality mapping-tables allowing for matching job vacancies and CVs in the European Job Mobility Portal. Specifically, we plan to achieve the following goals:… Leggi tutto (i) derive a machine-readable structure of the lexicon used within the Italian National Occupation Taxonomy (CP2011); (ii) connect ESCO to Italian National Taxonomy by means of word-embedding similarities; (iii) exploit on domain-experts to review and validate the results. It is also interesting to highlight that use of AI algorithms is well suited for natural language processing and reasoning as they allow deriving similarities between terms from a lexicon point of view. These similarities, in turn, can be used as a reference-table to match two distinct terms of taxonomies, making a bridge between taxonomies in a bottom up approach. The UNIMIB-CRISP team is composed by researchers and staff members that fits the Core competences of multidisciplinary approach to LMI. AI4ESCO will make a bridge between Italian Official Occupation Taxonomy towards ESCO (Occupation pillar) suggesting the correspondence between National occupation to ESCO (and vice-versa) in the job vacancy classification. We plan to evaluate the ability of AI4ESCO to provide the most relevant connection towards ESCO. The last impact is to discover lexicon similarities between ESCO occupations. The sustainability of AI4ESCO that will allow the mapping table to be updated over time will be guaranteed by: specific and reproducible report describing in detail the steps made to realise the AI system which could work as a manual for a future re-implementation of the process for other national taxonomies; code to reproduce and re-iterate the procedure that maps the Italian Occupation taxonomies on ESCO; use of Multi-criteria-decision-making model to provide evidences of the judgments and criteria used by labour market experts.