“Literature Review on Mitigating Bias in Algorithmic Systems”
The CyCAT team as part of the deliverables of the project, conduct a comprehensive survey of the literature on mitigating algorithmic bias. The survey developed a conceptual framework for understanding the problem and solution spaces of algorithmic bias, as well as the roles of various stakeholders. The manuscript was prepared as a submission to the journal ACM Computing Surveys. It provides a “fish-eye view” examining approaches across four areas of research: machine learning (ML), human-computer interaction (HCI), recommender systems (RecSys), and information retrieval (IR). The literature describes three steps toward a comprehensive treatment – bias detection, fairness and explainability management – and underscores the need to work from within the system as well as from the perspective of stakeholders in the broader context.
The goal of this STSE project will be to conduct a literature review and update the collected papers (Zotero repository) with more recent publications. As a first step, the researcher should collect recent publications related to the bias treatment (bias detection, fairness and explainability management) on the four aforementioned research areas. As a further step, the research communities can be extended by adding new ones or by focusing more on specific application domains e.g. criminal justice systems, recruitment systems.
Guidelines:
- A particular deliverable should result, e.g., a research paper, a white paper, a demonstration, etc.
- If travelling is allowed, CyCAT will support (via STSE) the travelling costs of the researcher to travel to Cyprus for presenting the research results at a CyCAT meeting. If travelling will not be allowed, due to the COVID-19 restrictions, the collaboration and the presentation of the results will be done remotely.