{"closest_prior_art":[{"name":"Aware: controlling false discoveries during interactive data exploration","overlap":"Identifies the same general failure mode: interactive tools expose users to many implicit hypotheses, increasing false discoveries, and integrates multiple-testing control into exploration.","remaining_difference":"It is not a library-collection workflow and does not disclose a cycle-level attempted-look registry, Holm-gated withdrawal status, untouched subsequent-period confirmation, branch-context review, or collection-action authority gate.","source_ids":["SRC4"]},{"name":"OCLC GreenGlass","overlap":"Provides interactive, multifaceted analysis of circulation, subjects, call numbers, holdings, and other attributes to construct withdrawal and preservation scenarios for low-use collections.","remaining_difference":"The retained product description does not disclose automatic enumeration of every attempted analysis, familywise adjustment based on that opportunity set, exploratory-status enforcement, or fresh-period confirmation before action eligibility.","source_ids":["SRC3"]},{"name":"ALA collection-maintenance and weeding guidance","overlap":"Treats circulation as one input to weeding, requires broader collection context and trained-librarian judgment, and describes multi-step approval in academic libraries.","remaining_difference":"It does not prescribe statistical treatment of dashboard multiplicity, an append-only claim registry, locked discovery/confirmation data boundaries, or a Holm rule.","source_ids":["SRC2"]},{"name":"White study of circulation and citation evidence for monograph deselection","overlap":"Empirically challenges circulation-only deselection by showing that borrowing history can omit significant titles and may not reliably predict future use.","remaining_difference":"It recommends additional evidence but does not address false discoveries caused by selectively reporting one result from many dashboard slices or prescribe registration and prospective confirmation.","source_ids":["SRC1"]}],"contrastive_claim_falsifier":"The contrast is falsified if an existing library product or documented routine practice is shown to automatically retain the complete set of attempted collection-use analyses, make that set determine a locked familywise status, prevent exploratory results from entering the action queue, and require a specification-locked test on untouched subsequent-period data before branch and authorized-director review; operationally, it is also falsified if the proposed shadow pilot cannot reconstruct looks or prevent unconfirmed leads from reaching that queue.","contrastive_claim_remaining":"For consequential collection claims selected through interactive circulation slicing, a cycle-level attempted-look register coupled to a predeclared Holm gate and untouched-period confirmation will provide an auditable action-eligibility barrier not supplied by circulation thresholds, multifactor deselection tools, professional contextual review, or generic interactive multiple-testing control alone.","experiment_id":"eoa_inverse_innovation_exp13_second_slot_policy60_20260806","gates":{"adequate_source_search":{"rationale":"Searches covered the proposal directly, older terms such as deselection and bibliomining, products and guidance, and component combinations involving interactive exploration, multiplicity, familywise control, registration, and holdout confirmation. Four direct sources from three organizational publishers were opened; they include primary research, official guidance, and a first-party product source. No exact library-specific implementation appeared, but that bounded miss is not treated as novelty evidence.","source_ids":["SRC1","SRC2","SRC3","SRC4"],"status":"PASS"},"bounded_next_test":{"rationale":"A read-only pilot limited to two branches, ten frozen classification bands, one discovery cycle, and one subsequent confirmation period has explicit inputs, stopping conditions, privacy suppression, process outcomes, and no authority to alter collections. It can test logging completeness, status enforcement, contamination, and queue leakage without requiring a full deployment.","source_ids":["SRC2","SRC3","SRC4"],"status":"PASS"},"distinct_testable_claim":{"rationale":"Interactive multiple-testing control is established adjacent art, while library tools and guidance already combine circulation with contextual criteria. The remaining claim is nevertheless distinct and falsifiable because it requires their specific workflow conjunction: exhaustive attempted-look registration, a locked Holm withdrawal gate, enforced labels, and untouched-period confirmation before decision eligibility.","source_ids":["SRC1","SRC2","SRC3","SRC4"],"status":"PASS"},"no_obvious_safety_or_authority_stop":{"rationale":"The proposed first step is advisory, read-only, aggregate, privacy-suppressed, and reversible; collection changes remain with the designated authority after branch review. This is consistent with guidance favoring trained professional judgment and multi-step approval rather than automated low-use removal. Privacy failure and premature action are explicit halt conditions.","source_ids":["SRC2","SRC3"],"status":"PASS"},"supported_problem":{"rationale":"The sources show that real collection systems support multifaceted, interactive circulation-based withdrawal analysis; ALA guidance confirms circulation is used in weeding while warning that low usage alone can mislead; empirical research finds borrowing history can omit significant works and poorly predict future use; and interactive-data research establishes that many implicit hypotheses can inflate false discoveries. The specific assertion that libraries routinely omit inspected alternatives from final packets was not directly documented, so support is partial rather than complete.","source_ids":["SRC1","SRC2","SRC3","SRC4"],"status":"PASS"}},"prior_art_disposition":"ADJACENT_PRIOR_ART","problem_evidence":{"finding":"The component conditions are visible: circulation influences consequential deselection, commercial tools permit multifaceted interactive slicing, circulation-only evidence has documented limitations, and interactive exploration creates multiple-testing risk. No retained source directly measures undocumented alternative looks or selected-slice reporting in an actual library review packet.","source_ids":["SRC1","SRC2","SRC3","SRC4"],"status":"PARTLY_SUPPORTED"},"research_id":"eoa_inverse_innovation_exp13_light_screen_20260806","schema_version":1,"screen_id":"E13P130","screen_survival":true,"search_lanes":{"component_combination":{"no_result_note":"No retained result disclosed the complete library-specific combination of attempted-look registration, Holm familywise gating, locked status labels, untouched subsequent-period confirmation, and collection-action authority review.","queries":["Holm correction holdout validation library circulation collection management","library weeding circulation analysis preregistration holdout confirmation multiple comparisons","collection management deselection familywise Holm circulation","interactive visual analytics multiple comparisons dashboard false discovery rate primary research"],"source_ids":["SRC3","SRC4"]},"direct_problem_and_intervention":{"no_result_note":"Direct searches found the underlying workflow and statistical problem separately, but no exact implementation of the proposed end-to-end gate.","queries":["library collection weeding circulation data dashboard low use multiple comparisons statistical significance","library collection assessment data mining circulation analysis cherry picking multiple testing","library deselection circulation analysis data dredging cherry picking hidden multiplicity"],"source_ids":["SRC1","SRC3","SRC4"]},"products_practices_and_standards":{"no_result_note":null,"queries":["library deselection weeding circulation statistics many criteria branches classification time period","library weeding decision support tool filters circulation branch classification GreenGlass OCLC","library deselection circulation analysis products practices standards"],"source_ids":["SRC1","SRC2","SRC3"]},"synonyms_and_historical_terms":{"no_result_note":"Terms including weeding, deselection, deaccessioning, bibliomining, data dredging, data snooping, cherry-picking, and hidden multiplicity produced adjacent practices and research but no exact library-specific match.","queries":["library deselection circulation analysis data dredging OR cherry picking OR hidden multiplicity","library bibliomining circulation trend analysis withdrawal decision audit trail","weeding deselection deaccessioning circulation analysis multiple comparisons"],"source_ids":["SRC1","SRC2","SRC4"]}},"sources":[{"claims_supported":["Circulation counts and date-of-last-use are heavily emphasized in data-driven deselection.","Circulation-only analysis can omit influential titles and assumes past borrowing predicts future use.","The study concludes that circulation and citation evidence should both inform research-monograph deselection."],"publisher":"College & Research Libraries, Association of College and Research Libraries","source_id":"SRC1","source_type":"PRIMARY_RESEARCH","title":"Citations and Circulation Counts: Data Sources for Monograph Deselection in Research Library Collections","url":"https://crl.acrl.org/index.php/crl/article/view/16567"},{"claims_supported":["Circulation history is an established weeding criterion.","Low usage should not override local importance, uniqueness, or professional context.","Trained librarians and, in academic settings, multi-step approval should participate in weeding decisions."],"publisher":"American Library Association","source_id":"SRC2","source_type":"OFFICIAL_GUIDANCE","title":"Collection Maintenance and Weeding","url":"https://www.ala.org/tools/challengesupport/selectionpolicytoolkit/weeding"},{"claims_supported":["GreenGlass is an interactive collection decision-support product.","It combines circulation and item data with holdings and authoritative lists.","Its multifaceted Query Builder constructs withdrawal and preservation scenarios and segments collections by several attributes."],"publisher":"OCLC","source_id":"SRC3","source_type":"FIRST_PARTY_PRODUCT","title":"GreenGlass: Evidence-based decision support for libraries","url":"https://www.oclc.org/en/greenglass.html"},{"claims_supported":["Interactive exploration exposes users to a large body of implicit hypotheses and can increase false discoveries.","Ordinary false-discovery procedures are not directly suited to adaptive interactive exploration.","The authors implemented and experimentally evaluated multiple-testing controls in an interactive system."],"publisher":"ACM; record hosted by the German Research Center for Artificial Intelligence (DFKI)","source_id":"SRC4","source_type":"PRIMARY_RESEARCH","title":"Controlling False Discoveries During Interactive Data Exploration","url":"https://www.dfki.de/web/forschung/projekte-publikationen/publikation/14459"}],"world_novelty_boundary":"This bounded public-web screen found adjacent statistical controls, library decision-support products, professional weeding guidance, and circulation-evidence research, but no exact retained match for the complete proposal. It cannot establish world novelty, patentability, market size, expert acceptance, implementation feasibility beyond the bounded test, or realized value."}