Pandemic Journaling Project
Distant reading on pandemic diaries - novelty, uncertainty, and feeling across the year.
Problem
During the COVID-19 pandemic, thousands of people contributed personal journal entries to the Pandemic Journaling Project, a repository organized by the University of Connecticut's Department of English. Read one entry at a time, the archive is thousands of individual stories - it can't show you the patterns that only exist across the whole pile.
System design & approach
Using Mathematica and Wolfram Language, I pulled the entire corpus into an analysis notebook, cleaned the text, split it into words and sentences, then ran part-of-speech filters, custom bigrams, and word-position plots. I focused on the context surrounding words to get at the collective human experience underneath individual entries - what the course called "distant reading": turning thousands of individual stories into insights no one could reach by reading them one at a time.
My role
Solo project - I built the entire analysis notebook and every visualization in it myself, from raw corpus to finished plots. The guided walkthrough below rebuilds the notebook's key steps, challenges and findings included - rank the most common adjectives before seeing the real answer, then scrub through the year to see feeling shift.
Walk the analysis
Before the analysis, a taste of the archive itself - real entries from the Pandemic Journaling Project archive, cycling below. Then let's walk through what the notebook found, step by step.
entry 01 of 10
“Today is day 87 since I began quarantine. Philadelphia entered yellow phase of re-entry. Nothing changed for me.”
from Philadelphia
For more entries check the Pandemic Journaling Project