Our whole world is driven by datas, informations, knowledges, we see them every day. The data mining methods help in many ways, because they can recognize patterns in datasets, can draw conclusions from the interrelations.
People are connected in many ways today. These connections can form a social network and people can share their informations through these networks. One of these informations is the actual location and Foursquare used this kind of share, to build their system.
The database of Foursquare contains a lot of valuable information about users and venues and of course about the connections between them.
If we see a city's structure, how dominates the different venue categories, it would be interesting in many aspects. The database of Foursquare is very useful for this question.
I would like to find the local outbrakes, so i need to make a density classification from the similar categories. The DBSCAN datamining algorithm is a good point to start.
I am focused only the dominant category by area, so i averaged the result of the clustering on a geographical grid. Now i have a good summary about what are the dominant categories by area.