Self-induced bias of recommender systems

Recommendation algorithms trained on a training set containing suboptimal decisions may increase the likelihood of making more bad decisions in the future. We call this harmful effect self-induced bias, to emphasize that the bias is driven directly by the user's past choices. In order to better understand the nature of self-induced bias of recommendation algorithms used by older adults with cognitive limitations, I have used agent-based simulation of e-commerce platform.

During the presentation, I will briefly introduce the most common recommender system types and explain the biases embedded in these algorithms. Then I will demonstrate my proposals for measuring and counteracting self-induced bias.