This interdisciplinary project aims at monitoring how occupational task and skill profiles are changing due to digitalization and at assessing the related consequences for workers’ career outcomes. To this end, the study will apply advanced text mining techniques to large and multilingual corpora of job ad texts, consisting of over 9 million texts, in order to build information extraction solutions for job tasks, skill requirements, occupations, and ICT tools. The information extraction pipeline will allow, first, the semantic indexing of job ads and, second, building a monitoring of tasks and skills reflecting employer demand. The resulting resources and methods will be shared with the computational social science community and advance the state-of-the-art in the field.
The semantic enrichment is indispensable for identifying trends in occupational task profiles and in demand for skill profiles, particularly whether they increasingly include ICT- and ICT-complementary tasks and skills, become more specialized or diversified and how skill profiles diverge or converge between occupations. The semantic enrichment is furthermore necessary for understanding how digitalization operates (e.g., within-occupation shift and between-occupation change) and for assessing its significance as a driver of these trends.
How the identified characteristics of changing task profiles and demand for skill profiles affect workers’ career outcomes will be examined with regard to unemployment, occupational change, wages and status mobility. Results will provide much needed evidence of how the digital transformation of the labor market generates career development opportunities for some workers and negatively affects others. Blending sociological expertise with computational linguistics, this project assesses the digital transformation and provides scientific evidence on how labor markets develop in the digital economy and what the consequences for workers are.
The first practical benefit of this study pertains to the setup of a public web-based data-mining dashboard for task and skill monitoring of the Swiss labor market. A workshop targeted to labor market stakeholders will help identify prevalent and common information needs regarding task and skill monitoring. Hence, more targeted data analytics pipelines could be developed in future spin-off projects with our strong implementation partner x28 Inc. Second and equally important, our implementation partner SECO will profit from the complete set of project resources for implementing skill-based matching of jobseekers and vacancies for the Swiss public employment service. Third, a further workshop will summon labor market stakeholders to inform them about the consequences of the digital transformation on workers’ careers. This evidence is prerequisite for scrutinizing curricula of professional training programs for their adequacy and for devising measures targeted to skill acquisition and retraining. Fourth, this study will also be in a position to potentially forecast skill demand as it combines time series data of job ads covering long time periods with a large job ad dataset ideal for in-depth analyses.
Simon Clematide (Leading NLP researcher)
Ann-Sophie Gnehm (External Doctoral Student)
This project is funded as part of the Swiss National Science Foundation's (SNSF) National Research Programme “Digital Transformation” (NRP 77, link). The programme's objective is to investigate the dynamics of the digital transformation in Switzerland.