mono2010/DSI_Client_Project

Scraped data from the South African government to locate and analyze low income neighborhoods # Improve Slum Area Identification through Real-Estate Data (problem 18) #### File Directory: - Assets (Contains shapefiles R scripts used to create final map.) - Code (Contains code used to scrape government website + clustering algorithm.) - Data (Raw data from scrape and population density data.) - Graphics (Geotiffs and images used in ReadMe + Power Point) - ReadMe.md (ReadMe) - city_final.html (Final Leaflet Interactive Map of City of Johannesburg Metropolitan Municipality) - .gitignore (Gitnore) ## Friday, February 21, 2020. git.generalassemb.ly ## Team members: |Michael Ono|Dylan Blough|Neil Hamlett| |---|---|---| ## Problem Statement. Mapping of ***informal settlements*** with satellite imagery is a long-standing practice, but such methods could be enhanced through web-scraped real-estate data. This project would build a web scraper to house and apartment adverts for a selected city in Africa/Latin America/Middle East. The scraper should download all adverts in the city during a recent period (ideally 3 years or more); and map all the adverts. The project should test the feasibility of estimating informal tenure from this information. Using gridded population estimates (e.g. from Facebook), the team would calculate the ratio of real estate adverts with population density. This ratio could serve as an input to machine learning models aimed at mapping informal settlements. The team selected Johannesburg, SA for purposes of prototyping. Johannesburg meets multiple criteria. It is a major population center in a developing economy. This is a client criterion. For data-availability purposes, SA is the most-developed economy on the Africa continent. ## Elaboration. ***Informal settlements*** (*aka* "slums") can be defined as " highly populated urban residential area consisting mostly of closely packed, decrepit housing units in a situation of deteriorated or incomplete infrastructure, inhabited primarily by …