Demographic Inference and Representative Population Estimates from Multilingual Social Media Data
This work addresses the challenge of achieving representative social sensing from multilingual social media data, which is crucial for downstream applications in global contexts, though it is incremental in improving existing inference and correction techniques.
The authors tackled the problem of non-representative and biased demographic inference from multilingual social media data by developing a multimodal deep neural architecture for joint classification of age, gender, and organization-status in 32 languages, which substantially outperforms current state-of-the-art methods while reducing algorithmic bias, and combined it with interpretable multilevel regression for bias correction to enable more accurate population estimates in a large experiment over European regions.
Social media provide access to behavioural data at an unprecedented scale and granularity. However, using these data to understand phenomena in a broader population is difficult due to their non-representativeness and the bias of statistical inference tools towards dominant languages and groups. While demographic attribute inference could be used to mitigate such bias, current techniques are almost entirely monolingual and fail to work in a global environment. We address these challenges by combining multilingual demographic inference with post-stratification to create a more representative population sample. To learn demographic attributes, we create a new multimodal deep neural architecture for joint classification of age, gender, and organization-status of social media users that operates in 32 languages. This method substantially outperforms current state of the art while also reducing algorithmic bias. To correct for sampling biases, we propose fully interpretable multilevel regression methods that estimate inclusion probabilities from inferred joint population counts and ground-truth population counts. In a large experiment over multilingual heterogeneous European regions, we show that our demographic inference and bias correction together allow for more accurate estimates of populations and make a significant step towards representative social sensing in downstream applications with multilingual social media.