I often browse Science Daily in the morning. It's an aggregation site for science and technology news. It does a good job of providing links for further information, which I appreciate, since publications sometimes mangle reporting on research, and direct access to the study itself can be valuable.
This morning I read an article, posted on Science Daily in late December, about an interesting use of geolocalized tweets to extract information about existing use of land in urban areas. This data is useful in urban planning, because it gives a good sense of what large groups of people are doing from hour to hour in a specific city or region. The article mentions night life in particular, where the data might give guidance into allocation of resources for cleanup and noise mitigation.
The study was conducted by two Spanish computer scientists, and published in the journal Engineering Applications of Artificial Intelligence. It used data from Madrid, Manhattan, and London.
I have a long-standing interest in urban planning, and the data and analysis featured in the study could clearly provide guidance in encouraging a more even mix of uses in a city and region, between such types as residential, commercial, industrial, and entertainment.
Like most activities involving geolocation and data mining, it could also have a downside, undermining privacy, and enabling corporate or governmental social engineering. I have not explored the darker side of this particular use of technology to my own satisfaction, but the article is worth reading, and I look forward to slogging through the study itself.