Limitations on Robust Ratings and Predictions - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2016

Limitations on Robust Ratings and Predictions


Predictions are a well-studied form of ratings. Their objective nature allows a rigourous analysis. A problem is that there are attacks on prediction systems and rating systems. These attacks decrease the usefulness of the predictions. Attackers may ignore the incentives in the system, so we may not rely on these to protect ourselves. The user must block attackers, ideally before the attackers introduce too much misinformation. We formally axiomatically define robustness as the property that no rater can introduce too much misinformation. We formally prove that notions of robustness come at the expense of other desirable properties, such as the lack of bias or effectiveness. We also show that there do exist trade-offs between the different properties, allowing a prediction system with limited robustness, limited bias and limited effectiveness.
Fichier principal
Vignette du fichier
428098_1_En_8_Chapter.pdf (323.18 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01438352 , version 1 (17-01-2017)





Tim Muller, Yang Liu, Jie Zhang. Limitations on Robust Ratings and Predictions. 10th IFIP International Conference on Trust Management (TM), Jul 2016, Darmstadt, Germany. pp.113-126, ⟨10.1007/978-3-319-41354-9_8⟩. ⟨hal-01438352⟩
46 View
67 Download



Gmail Facebook X LinkedIn More