PRISM: Progressive Reduction and Inference for Structured Multi-observer Document Classification with Absent Implicit Signals
Fuente:
Zenodo
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Lingua: | inglese |
| Pubblicazione: |
Zenodo
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866901184557613056 |
|---|---|
| author | Jaime, Alejandro |
| author_facet | Jaime, Alejandro |
| contents | <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">Document classification in enterprise settings faces a fundamental challenge that the literature has not formally addressed: documents where the discriminative signal is not concentrated in any fixed-size context window but distributed across the full document content. We call this class Documents with Absent Implicit Classification (DAIC). Existing approaches — fine-tuned BERT variants, Longformer, and direct LLM classifiers — achieve only 40–73% accuracy on DAIC documents.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">We propose PRISM (Progressive Reduction and Inference for Structured Multi-observer classification), a multi-observer architecture that reframes document classification as progressive evidence accumulation and hypothesis space reduction. Five specialised filters (F0–F4) observe the same document from complementary perspectives — each a tile of evidence — reducing the candidate type set while satisfying four formal properties: monotonicity, correctness, convergence, and epistemic honesty. The final LLM Arbiter reasons over approximately 1.8 candidates with accumulated evidence, not over the full type space.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">We evaluate PRISM on a real-world corpus of 2,037 administrative inspection documents (Admin-Legal corpus, anonymised) across 14 document types. Approximately 40% of the corpus consists of DAIC-genuine documents with no explicit type declaration; the remaining 60% carry an explicit type marker and are trivially classifiable. PRISM achieves 91.25% overall accuracy on the complete test set and 90% accuracy on the DAIC-genuine subset, as measured by two independent domain-expert validators external to model development. Median inference latency is 1.0ms; only 2.8% of documents reach the LLM Arbiter.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">PRISM requires no manual annotation, no predefined type schema, and no prompt engineering — it learns exclusively from operational folder structure, inheriting tacit professional knowledge accumulated by domain experts over years of daily practice.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">This paper makes seven contributions: (1) formal definition of the DAIC problem and characterization of why existing approaches fail; (2) the PRISM architecture with proof of four formal properties; (3) the Document Dictionary Representation (DDR), a structured deduplicated view eliminating token repetition in long document encoding; (4) demonstration that an LLM Arbiter over a small candidate set substantially outperforms the same LLM over the full type space; (5) human-validated accuracy of 90% on the DAIC-genuine subset — to our knowledge the first reported IDP result on long-document classification under independent domain-expert validation on the hardest subset; (6) an automated ground truth auditing procedure that detected 27.9% label contamination from operational misfiling; (7) comprehensive ablation studies across all five filters.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19948470 |
| institution | Zenodo |
| language | eng |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | PRISM: Progressive Reduction and Inference for Structured Multi-observer Document Classification with Absent Implicit Signals Jaime, Alejandro document classification, long document understanding, intelligent document processing, IDP, multi-observer architecture, hypothesis space reduction, LLM arbitration, epistemic honesty, auditable AI, DAIC, cascaded classification, enterprise document intelligence, label noise, ground truth auditing <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">Document classification in enterprise settings faces a fundamental challenge that the literature has not formally addressed: documents where the discriminative signal is not concentrated in any fixed-size context window but distributed across the full document content. We call this class Documents with Absent Implicit Classification (DAIC). Existing approaches — fine-tuned BERT variants, Longformer, and direct LLM classifiers — achieve only 40–73% accuracy on DAIC documents.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">We propose PRISM (Progressive Reduction and Inference for Structured Multi-observer classification), a multi-observer architecture that reframes document classification as progressive evidence accumulation and hypothesis space reduction. Five specialised filters (F0–F4) observe the same document from complementary perspectives — each a tile of evidence — reducing the candidate type set while satisfying four formal properties: monotonicity, correctness, convergence, and epistemic honesty. The final LLM Arbiter reasons over approximately 1.8 candidates with accumulated evidence, not over the full type space.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">We evaluate PRISM on a real-world corpus of 2,037 administrative inspection documents (Admin-Legal corpus, anonymised) across 14 document types. Approximately 40% of the corpus consists of DAIC-genuine documents with no explicit type declaration; the remaining 60% carry an explicit type marker and are trivially classifiable. PRISM achieves 91.25% overall accuracy on the complete test set and 90% accuracy on the DAIC-genuine subset, as measured by two independent domain-expert validators external to model development. Median inference latency is 1.0ms; only 2.8% of documents reach the LLM Arbiter.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">PRISM requires no manual annotation, no predefined type schema, and no prompt engineering — it learns exclusively from operational folder structure, inheriting tacit professional knowledge accumulated by domain experts over years of daily practice.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">This paper makes seven contributions: (1) formal definition of the DAIC problem and characterization of why existing approaches fail; (2) the PRISM architecture with proof of four formal properties; (3) the Document Dictionary Representation (DDR), a structured deduplicated view eliminating token repetition in long document encoding; (4) demonstration that an LLM Arbiter over a small candidate set substantially outperforms the same LLM over the full type space; (5) human-validated accuracy of 90% on the DAIC-genuine subset — to our knowledge the first reported IDP result on long-document classification under independent domain-expert validation on the hardest subset; (6) an automated ground truth auditing procedure that detected 27.9% label contamination from operational misfiling; (7) comprehensive ablation studies across all five filters.</p> |
| title | PRISM: Progressive Reduction and Inference for Structured Multi-observer Document Classification with Absent Implicit Signals |
| topic | document classification, long document understanding, intelligent document processing, IDP, multi-observer architecture, hypothesis space reduction, LLM arbitration, epistemic honesty, auditable AI, DAIC, cascaded classification, enterprise document intelligence, label noise, ground truth auditing |
| url | https://doi.org/10.5281/zenodo.19948470 |