Laboratory: "data science for economics: job postings analysis and measuring ai usage"

Retrieve real-world job posting data from web platforms using both traditional scraping (Selenium) and AI-based scraping (ScrapeGraphAI).
Clean, transform, and organize complex textual datasets in a cloud environment using AWS services.
Build and train a binary classification model capable of identifying whether a job posting has been written or assisted by AI.
Apply topic modelling techniques to analyse linguistic patterns, thematic content, and differences between pre-AI and post-AI job postings.
Develop and present analytical results through interactive dashboards using Tableau.

At the end of the laboratory, students will be able to:
Acquire and collect real-world job posting data through both traditional and AI-assisted web scraping methods.
Clean, preprocess, and organize unstructured textual datasets, and manage them within a cloud-based data lake environment.
Apply Natural Language Processing techniques to build and evaluate a binary classification model capable of detecting AI-generated or AI-assisted job postings.
Perform topic modelling to extract and interpret thematic patterns in job postings and analyse changes in content, structure, and language over time.
Develop data visualizations and interactive dashboards to communicate analytical findings effectively to academic or professional audiences.
Critically assess the implications of AI adoption in the labour market and understand how data-driven methods can support research in labour economics and digital transformation.