About Livehoods
Livehoods offer a new way to conceptualize the dynamics, structure, and character of a city by analyzing the social media its residents generate.
With Livehoods, people can investigate and explore the factors that come together to shape the social dynamics of a city, including municipal borders, demographics, economic development, resources, geography, and architecture.
With Livehoods, people can investigate and explore the factors that come together to shape the social dynamics of a city, including municipal borders, demographics, economic development, resources, geography, and architecture.
Data
The team combined approximately 11 million foursquare check-ins from the dataset of Chen et al. (2011) with their own dataset of 7 million check-ins that they downloaded between June and December of 2011. Foursquare check-ins are by default not publicly visible on social networks such as Twitter. These 18 million check-ins were all collected from the Twitter public timeline, then were aligned with venue information from the foursquare API. For each check-in, their data consists of the user ID, the time, the latitude and longitude, the name of the venue, and the category of the place.
Analysis methods
- Clustering Algorithm: The team present a spectral clustering approach to the discovery of local urban areas from geospatial check-in data. Viewed as a graph, they connect each venue node with an undirected edge to its m nearest neighbors by geographic distance, and they weigh the edges according to the cosine similarity of the distributions of check-ins at the two venues. Then, they applied the variation of spectral clustering on the graph.
- Related clusters: They also developed a way to compare different clusters based on the similarity of the distributions of users that visit them. Again, they used a cosine similarity measure.