Moving Out

“An analysis of real estate behavior during pandemic COVID-19”

What it is about

Impacts of Coronavirus are everywhere. In-town migrations during COVID-19 pandemic were observed due to lockdown policies, economic slowdown, increased unemployment rate and safety concerns. Such moving trend can be reflected in property data indirectly. We used real-estate big data to explore trends in people's behavior over time under epidemic context. The results could give people insights into location pattern changes and price variations in the real estate offer.

How we built it

After we obtain big real estate datasets of the five countries in South America, we started data pre-processing on Python. First, we reorganized data structures and determining capital cities as study areas based on record distributions. With further filtering, we decided to conduct analysis for rental and selling market respectively, and only focus on two property types: housing and apartment. A geospatial platform named Unfolded was then used for data visualization. Finally, time-series animations on rental and selling markets offer for the five countries were generated.

Challenges we ran into

The datasets used included detailed information which made for us to take some time to narrow down the scope and filter out valid information. Furthermore, deciding on a proper visualization method was also a challenge for us at first.

What we're proud of

We involved spatiotemporal analysis in the project, observing the spatial patterns over time. Besides, the outputs on Unfolded platform are highly interactive: the user can interact with datasets by defined functions or formats.

What we learned

Big data is a powerful way to understand market behavior. By properly processing and interpolating the datasets, some underlying patterns can be revealed. Besides, an appropriate visualization method is essential to an attractive and readable mapping product.

What's next

Reviews and comments would be collected to improve map design and data visualization to make the information of the real estate market offer easier to understand.

Sources

Sources We want to thank Properati (https://www.properati.com.ec/data), a real estate company operating in several countries around South America, for opening their data for this analysis.
Students
Jiaying Xue
Joel Salazar

10th intake
Supervisors
Juliane Cron, M.Sc.
Dr.-Ing. Mathias Jahnke
Keywords
COVID-19, Big Data, Real Estate, Location patterns
Try it