Determine appropriate protocol parameter choices

Determine appropriate parameter choices for the probabilistic focal-point detection protocol, including pseudonym-space and Bloom-filter parameters, as a function of the expected service-composition graph size while balancing correctness probability against computational feasibility.

Background

The proposed detection protocol uses blindable pseudonyms and Bloom filters. Its reliability depends on selecting a sufficiently large pseudonym space to make accidental pseudonym collisions negligible and configuring the Bloom filter to limit false positives, since a false-positive membership test can prematurely terminate recursive exploration and cause focal points to remain undetected.

The paper leaves the parameter-selection discussion unresolved and indicates that the appropriate values should depend on the expected size of the service-composition graph, potentially estimated through market analyses. The unresolved task is therefore to derive or validate parameter-selection rules that achieve an acceptable correctness probability without making the protocol computationally infeasible. This is relevant to deploying the protocol beyond the idealized analysis, which assumes that pseudonym collisions do not occur and Bloom-filter checks are correct.

References

We defer the discussion of choosing these parameters to a future work as we propose they should depend on the expected graph size (which could be derived from market analyses) and balance correctness with the aspects of computational feasibility.

— Identification of Compositional Risks in Data Protection Impact Assessments and Beyond  (2609.01201 - Graßhoff et al., 1 Sep 2026) in Section 4.2, “Threat Model and Parameter Choice” (subsubsection labeled subsubsec: threat model)