Abstract
This paper is a conceptual prior-art record, not a product and not a method. It records that an
automated system could infer the categories named below from interview or session audio, video,
or conversation, from a covert in-session task, or from a reciprocal bait story, and could use that
inference to rank, filter, hide, price, route, or exclude a person. Hiring and workplace use, for
applicants and existing employees, is the primary example. The same concept is recorded, as a
line item only, for other application domains. One legal note is given: such systems are, at the
date of this draft, unlawful in many jurisdictions. Protective counterparts are cross-referenced to a
companion disclosure. This version adds a generic, architecture-level description of an inference
stage feeding a decision stage, and of score fusion or weighting of inferred attributes into a
ranking, gating, or pricing output, with the categories of output across the listed domains.
Enablement of that pipeline is by citation to published, generic attribute-classifier, emotion-
recognition, and multimodal-fusion literature, not by a cue list or a training recipe. This paper
names categories and domains only. It supplies no physical or biometric cues, no detection
features, no models, and no thresholds for detecting any characteristic, and no scripts, prompts, or
bait lines.
Keywords: inferred protected characteristics; automated exclusion; hiring; workplace; ranking;
filtering; conceptual prior art.
Creative Commons License

This work is licensed under a Creative Commons Attribution-Share Alike 4.0 License.
Recommended Citation
Boll, Joe, "Automated Exclusion Based on Inferred Protected Characteristics: A Conceptual Prior-Art Record", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/13038