Library Genesis is one of the oldest and largest illegal scholarly book collections online. Without the authorization of copyright holders, this shadow library hosts and makes more than 2 million scholarly publications, monographs, and textbooks available. This paper analyzes a set of weblogs of one of the Library Genesis mirrors, provided to us by one of the service’s administrators. We reconstruct the social and economic factors that drive the global and European demand for illicit scholarly literature. In particular, we test if lower income regions can compensate for the shortcomings in legal access infrastructures by more intensive use of illicit open resources. We found that while richer regions are the most intensive users of shadow libraries, poorer regions face structural limitations that prevent them from fully capitalizing on freely accessible knowledge. We discuss these findings in the wider context of open access publishing, and point out that open access knowledge, if not met with proper knowledge absorption infrastructures, has limited usefulness in addressing knowledge access and production inequalities.

No abstract

Field experimentation has become a well-established practice to estimate individual treatment effects. In recent years, the Active Learning (AL) literature has developed methods to optimize the design of field experiments and reduce their cost. In this paper, we propose a novel AL algorithm for individual treatment effect estimation that works in batch mode for cases where the outcomes of an intervention are not immediate. It uniquely combines Expected Model Change Maximization and Bayesian Additive Regression Trees. Our approach (B-EMCMITE) uses the predictive uncertainty around the individual treatment effects to actively sample new units for experimentation and decide which treatment they will receive. We perform extensive simulations and test our approach on semi-synthetic, real-life data. B-EMCMITE outperforms alternative approaches and substantially reduces the number of observations needed to estimate individual treatment effects compared to A/B tests.