Causal Discovery as Dialectical Aggregation: A Quantitative Argumentation Framework
For researchers in causal discovery, QACD offers a more robust alternative to brittle constraint-based methods in finite-sample settings, though improvements are incremental.
Constraint-based causal discovery is brittle in finite samples due to cascading errors from conditional-independence tests. QACD represents CI outcomes as graded arguments and aggregates them via witness propagation, improving structural coherence and interventional reliability in noisy regimes.
Constraint-based causal discovery is brittle in finite-sample regimes because erroneous conditional-independence (CI) decisions can cascade into substantial structural errors. We propose Quantitative Argumentation for Causal Discovery (QACD), a semantics-driven framework that represents CI outcomes as graded, defeasible arguments rather than irreversible constraints. QACD maps statistical test outcomes to argument strengths and aggregates conflicting evidence through connectivity-mediated witness propagation, producing a fixed-point acceptability labeling over candidate adjacencies. Experiments on standard benchmark Bayesian networks suggest that QACD improves structural coherence and interventional reliability in several noisy or inconsistent CI regimes, while remaining competitive with classical constraint-based, hybrid, and prior argumentation-based baselines.