anitera/CounterEval

CounterEval提供了一个包含30个反事实场景的人工评估数据集，旨在作为评估各种反事实解释生成框架的基准。这些场景基于知名数据集如Adult和Pima Indians Diabetes数据集，并添加了额外的人工场景以探索极端值。每个场景由206名通过Prolific平台招募的个体进行评估，评估指标包括可行性、一致性、信任、完整性、公平性和复杂性等。

CounterEval offers a human-evaluated dataset of 30 counterfactual scenarios, engineered to serve as a benchmark for evaluating a wide range of counterfactual explanation generation frameworks. The scenarios are based on well-known datasets like the Adult dataset and the Pima Indians Diabetes dataset, with additional fabricated scenarios to explore extreme values. Each scenario has been evaluated by 206 individuals recruited through the Prolific platform, across metrics such as Feasibility, Consistency, Trust, Completeness, Fairness, and Complexity.