DADIT: A Dataset for Demographic Classification of Italian Twitter Users and a Comparison of Prediction Methods

TL;DR

DADIT provides a large-scale, GDPR-compliant Italian Twitter dataset with gender, age, and location labels, enabling robust demographic prediction. The study demonstrates that incorporating user tweets alongside bios markedly improves gender and especially age prediction, with XLM-based fine-tuning delivering the strongest results and generalizing to German data. Scaling to larger encoders and adding extra user-level features yields further gains, while a simple ensemble can boost gender accuracy. The work underscores the value of text-rich, demographically annotated social media data for computational social science, while addressing privacy and representativeness considerations.

Abstract

Social scientists increasingly use demographically stratified social media data to study the attitudes, beliefs, and behavior of the general public. To facilitate such analyses, we construct, validate, and release publicly the representative DADIT dataset of 30M tweets of 20k Italian Twitter users, along with their bios and profile pictures. We enrich the user data with high-quality labels for gender, age, and location. DADIT enables us to train and compare the performance of various state-of-the-art models for the prediction of the gender and age of social media users. In particular, we investigate if tweets contain valuable information for the task, since popular classifiers like M3 don't leverage them. Our best XLM-based classifier improves upon the commonly used competitor M3 by up to 53% F1. Especially for age prediction, classifiers profit from including tweets as features. We also confirm these findings on a German test set.

Figures (2)

• Figure 1: F1 score and prediction coverage tradeoff.
• Figure 2: Representativness of Twitter users in DADIT.