Exploratory Data Project

Finding Signal in Events and Public Attention

This project investigates how real-world events propagate into measurable attention patterns. It is useful as a portfolio example because it shows disciplined ingestion, schema-aware processing, reusable feature design, and analytical framing that can also be applied to operations, product launches, incidents, and market-facing systems.

Core Question

When notable events occur, how quickly and strongly does public attention respond, and which event shapes create durable attention versus short-lived spikes?

Data Inputs

• USGS seismic events: Objective event records with clear timing and magnitude characteristics.
• Wikimedia pageviews: Aggregate public traffic patterns used as an attention signal.

Method

• Ingest each source independently with explicit schema contracts and repeatable processing logic.
• Define pre-event and post-event windows to compare attention behavior against a baseline.
• Measure lag-to-peak, peak magnitude, and decay duration as reusable analytical features.
• Favor explainable metrics over speculative modeling claims.

Why It Matters

The same design pattern can support event-response analysis in many business settings: outage monitoring, product launch analysis, customer support surges, and public response to operational incidents. It demonstrates careful data engineering, not just an interesting chart.