Smelly London: visualising historical smells.
summary
Smelly London brings a dataset of 5,500 MOH reports into reproducible analysis and visualization, mining OCR text with Python 3, NLTK, SQLite and multiprocessing to extract smell-related sentences, categorize terms and reduce processing from ~15 to ~5 minutes. It produces interactive maps with Leaflet and pie charts plus a time slider, links records to original reports, is building a nineteenth-century placegazetteer and ontology, and reports the maps and filters functioning though some web performance and historical-geocoding gaps remain.
AI-written summary · about
chapters (9)
- 0:00Smelly London project
- 3:53Medical Officer of Health reports
- 5:18Historical smells in London
- 9:10Data mining smells
- 11:39First smell map
- 14:58Advanced smell map
- 16:27Future plans for ontology
- 21:27Q&A on smell extraction
- 22:10Linking smells to health data
related talks
18mSmell Pittsburgh: community-empowered mobile smell reporting system
26mThe Data I Breathe
22mUnderstanding our cities with FOSS4G
35mVisualising Public Service: Seeing Inside Birmingham City
48mStreet Fighting Trend Research
42mTackling Homelessness with Open Data
01h 29mImproving Open Data Quality using Python
55mSocial Listening for Richer Data, Unstructured Analytics