Analysis of Ecological Inference methods in social network data

• Gustavo Penha Federal University of Minas Gerais / Hekima
• Thiago N. C. Cardoso Hekima
• Ana Paula Couto da Silva Federal University of Minas Gerais
• Mirella M. Moro Federal University of Minas Gerais https://orcid.org/0000-0002-0545-2001

DOI: https://doi.org/10.5753/sbbd.2016.24313

Abstract

Online Social Networks have recently become extremely popular and generate a huge volume of spontaneous data. Knowing demographics of these users can provide useful information (e.g. marketing campaign segmentation). Unlike most approaches, we propose the use of Ecological Inference for understanding demographics of groups of people. Our results show that it is possible to infer gender and age using only a census and aggregated information obtained from a social network such as aggregated support for a political candidate.

Keywords: Ecological inference, Demographic characteristics

References

Bi et al., B. (2013). Inferring the demographics of search users: social data meets search queries. In WWW.

Flaxman et al., S. R. (2015). Who supported obama in 2012?: Ecological inference through distribution regression. In SIGKDD.

Goodman, L. A. Some alternatives to ecological correlation. American Journal of Sociology.

Imai, K., Lu, Y., and Strauss, A. (2008). Bayesian and likelihood inference for 2×2 ecological tables: an incomplete-data approach. Political Analysis.

Jain, R. (2008). The art of computer systems performance analysis. John Wiley & Sons.

King, G. (1997). A solution to the ecological inference problem. Princeton, NJ: Princeton University Press.

King, G., Tanner, M. A., and Rosen, O. (2004). Ecological inference: New methodological strategies. Cambridge University Press.

Tumitan, D. and Becker, K. (2013). Tracking sentiment evolution on user-generated content: A case study on the brazilian political scene. In SBBD.

Wakefield, J. (2004). Ecological inference for 2×2 tables (with discussion). Journal of the Royal Statistical Society: Series A (Statistics in Society).

Willmott, C. and Matsuura, K. Advantages of the mean absolute error (mae) over the root mean square error (rmse) in assessing average model performance. Climate research.

Zhong, Y., Yuan, N. J., Zhong, W., Zhang, F., and Xie, X. (2015). You are where you go: Inferring demographic attributes from location check-ins. In WSDM.