Skip to main content

The PeerRank Method

Toby Walsh ( University of New South Wales )


I propose the PeerRank method for peer assessment. This constructs a grade for an agent based on the grades proposed by the agents evaluating the agent. Since the grade of an agent is a measure of their ability to grade correctly, the PeerRank method weights grades by the grades of the grading agent. The PeerRank method also provides an incentive for agents to grade correctly. It rewards agents who grade well, and penalises those that grade poorly. As the grades of an agent depend on the grades of the grading agents, and as these grades themselves depend on the grades of other agents, I define the PeerRank method by a fixed point equation similar to the PageRank method for ranking web-pages. I identify some formal properties of the PeerRank method, discuss some examples, compare with related work and evaluate the performance on some synthetic data.

Speaker bio

Toby Walsh is a professor of artificial intelligence at NICTA and the University of New South Wales. He has served as Scientific Director of NICTA, Australia's centre of excellence for ICT research. He has also held research positions in England, Scotland, Ireland, France, Italy, Sweden and Australia. He has been Editor-in-Chief of the Journal of Artificial Intelligence Research, and of AI Communications. He is Editor of the Handbook of Constraint Programming, and of the Handbook of Satisfiability.

Share this: