On 21/04/2017, Dr. Haishan Wu from Baidu Big Data Lab came to HKUST and gave an extremely interesting talk on spatio-temporal data mining on socio-economic factors. Following is a brief list of the works discussed in Dr. Wu's talk. Note that most of such works are coming from SCIENCE !

(1) Eagle, N., Macy, M. and Claxton, R., 2010. Network diversity and economic development. Science, 328(5981), pp.1029-1031.

This paper defines a metric called 'network diversity', and find that it is correlated to population-level socio-economics factors. [Data: 2005.08, mobile phone call logs in UK]

(2) Jean, N., Burke, M., Xie, M., Davis, W.M., Lobell, D.B. and Ermon, S., 2016. Combining satellite imagery and machine learning to predict poverty. Science, 353(6301), pp.790-794.

This paper uses nightlights as a proxy for economic activity to learn features from satellite daytime images, and then use these features to predict poverty level of countries. This is some kind of transfer learning and one of the best things for this paper is the satellite images are publicly available data (from Google Static Maps). Code for this paper can be found in https://github.com/nealjean/predicting-poverty.

(3) Chi, G., Liu, Y., Wu, Z. and Wu, H., 2015. Ghost Cities Analysis Based on Positioning Data in China. arXiv preprint arXiv:1510.08505.

This work is from Dr. Wu's group and was reported by MIT Technology Review. It uses Baidu query data to quantify the empty ratio of houses in different China cities (or more fine-grained districts). The system can be visited here: http://bdl.baidu.com/ghostcity/

(4) Blumenstock, J., Cadamuro, G., & On, R. (2015). Predicting Poverty and Wealth from Mobile Phone Metadata. Science, 350(6264), 1073-1076.

There is also one review paper about big data and economics from Science:

Einav, L. and Levin, J., 2014. Economics in the age of big data. Science, 346(6210), p.1243089.