Publication and Patent Data Collection

This lab matters because credible forecasts require transparent, reproducible data pipelines, not ad hoc searches. Building directly on Session 1's scoping work, students translate research questions into database queries and patent classifications. The session connects evidence sourcing to later network and NLP analyses by enforcing metadata discipline from day one. Students produce a search protocol and export log that documents query logic, filters, and coverage. Skills developed include query design, database navigation, and basic data validation. These outputs become the raw material for cleaning, deduplication, and analysis in the next session.