Airbnb Monitor
Developing a user-friendly, open-source dashboard for community organizers navigating complex geospatial data streams, enabling better insight into a shadowy network of illegal corporate hotels.
Project management
Community Engagement
Data mining + analytics
Web development
Dashboard design
Objectives
Data Mining and Analysis: Predict similarities and anomalies across numerous collections of geospatial data over time to predict violations of Boston's short-term rental ordinance.
User-Friendly Interface: Develop a user-friendly interface that offers meaningful access to public housing data, facilitating efficient data collection.
Community Empowerment: Enable community organizers to identify and address violations within a complex network of illegal corporate hotels impacting affordable housing.
Approach
Predictive Analytics: Building and implementing a custom machine learning clustering algorithm to infer likely violations of the ordinance.
Community Collaboration: Dashboard designed in collaboration with community leaders through testing and feedback, ensuring it aligns with the specific needs and workflows of organizers.
Technology Stack: Hosting the dashboard with Flask and PythonAnywhere for a robust and scalable solution.
Outcomes
Open-Source Dashboard: Launch an open-source JS+Python online dashboard, providing valuable insights to community organizers monitoring the local short-term rental market in Boston.
Usability Features: Offering citywide sorting, various filter options, and search functionality, complete with direct links to sources and detailed summaries.
Impactful Advocacy: Provided materials and analyses that supported City Council efforts to reform relevant legislation, showcasing the dashboard's practical application in driving practical benefits to local communities.
Data Lake
Custom-built predictive matching algorithms formed the “data lake” behind the dashboard, merging historic Airbnb listings, public housing data, the city’s official short-term rental license registry, and other geospatial reference data.
Partners and stakeholders from the city government and local community organizations provided critical feedback throughout the data collection and processing stages to ensure the accuracy and validity.
Design and features
At-a-glance citywide statistics
Based on feedback and interviews with various user groups, I developed an online dashboard presenting actionable statistics on the short-term rental market for use by community advocates.
Automated centroid analysis
To provide leaders with deeper insights on the short-term rental market in their neighborhood, my program monitors similar datapoints within a variable geographic radius, inferring the building or buildings containing a high number of illegal corporate Airbnb listings.
Detailed source references
The dashboard provides users with a list of matches for a given cluster, providing key source materials to support legislative reform efforts.
Advanced search and filter functions
The dashboard provides users with a list of matches for a given cluster, providing key source materials to support legislative reform efforts.
Forest Seminar
Leah Hager Cohen