Joanna Biega

4Machine learning papers tracked by Lacuna
Is this you?

Research Overview

Joanna Biega focuses on information retrieval, algorithmic fairness, and privacy in online communities. Her research trajectory spans from information extraction architectures for large-scale ontologies to assessing privacy risks emerging from user-generated text. Most recently, her work investigates how ranking algorithms and choice architectures impact the equitable distribution of attention and resources in socio-technical systems.

Research Trajectory

Joanna Biega's early research focused on knowledge bases and information extraction methodologies. In 2013, as first author, she presented Inside YAGO2s: a transparent information extraction architecture, detailing a modular framework for the YAGO ontology. This system allowed for the parallel extraction, verification, and merging of facts from diverse sources, providing a transparent interface for tracing data provenance and deduction rules.

The focus of her work subsequently shifted toward information retrieval (IR) and user privacy in online communities. In R-Susceptibility: An IR-Centric Approach to Assessing Privacy Risks for Users in Online Communities (2016), Biega led the development of a ranking-based methodology to evaluate privacy risks stemming from user-generated text. This research addressed the limitations of structured data privacy models by quantifying the exposure of users discussing sensitive topics—such as mental health—to potential adversarial monitoring.

By 2018, Biega's research expanded into algorithmic fairness, specifically addressing biases in ranking systems used for selection and recommendation. As first author of Equity of Attention: Amortizing Individual Fairness in Rankings, she investigated how position bias disproportionately limits the attention received by low-ranked subjects in environments like employment sites and the sharing economy. The paper proposed novel mechanisms to quantify and mitigate this unfairness by amortizing individual fairness across multiple ranking instances, aiming to ensure a more equitable distribution of opportunities over time.

Current Directions

The most recent work in the evidence, dating to 2019, investigates the intersection of platform design, human behavior, and societal inequality. Expanding on earlier themes of algorithmic fairness, this direction examines how interface design choices influence user decisions and resource allocation. As a co-author of On the Impact of Choice Architectures on Inequality in Online Donation Platforms, Biega contributed to research analyzing socio-technical systems like crowdfunding and donation sites. This work highlights how choice architectures—the environments in which funding decisions are made—can either counterbalance or inadvertently exacerbate the unfair distribution of financial support to entities in need.

Profile updated

See something wrong? Report a mistake