A production extraction layer that converts a live document stream into a validated knowledge graph aligned to a formal ontology, and improved search recall from roughly 70 to 95 percent with no false merges, and corrected seven classes of silent quality defect.
Abstract
Large language models extract entities and relationships from unstructured documents fluently but inconsistently: type vocabularies fracture across documents, the same person surfaces under several name variants, relationships duplicate, and distinct individuals who share a name risk silent conflation. This paper presents the design, implementation, and empirical refinement of a production extraction layer that converts a live document stream into a validated knowledge graph aligned to a formal ontology. The system consumes document metadata from Kafka, routes PDF, spreadsheet, Office, and image content through handlers built for each format, and extracts entities and relationships in two passes using a locally hosted Qwen3.5-9B model tuned on the ontology. Its distinguishing component is ontology-guided extraction: the relevant slice of a curated ontology is retrieved live from a graph database by embedding similarity and injected into the extraction prompt, reducing catalog overhead by about 94 percent relative to static domain slices. Extracted results then pass through a refinement pipeline of five stages: deterministic cleaning, merging across chunks, a second pass for relationships, six deduplication algorithms that require no model inference, and an embedding resolution subsystem whose conflict guard no similarity score can override. Evaluation on intelligence corpora improved search recall from roughly 70 to 95 percent with no false merges, and corrected seven classes of silent quality defect, ranging from a bug that truncated source text by a single character to the systematic duplication of entities that carried title prefixes.
This work targets a KG for Sophocles’ Antigone that supports two coupled uses: structured retrieval, through integrity and competency questions expressed in SPARQL over dramatic structure and interpretive annotations; and interactive exploration, through a lightweight read client that navigates lines across languages, shows scene context, and reports corpus statistics.
Extraction produces candidate entities and relationships; writing them into a graph is where identity is decided, and identity decisions are destructive in a way extraction errors are not. A wrong type can be corrected later, but two records merged under one identity cannot be separated once their properties have been combined, and the merge leaves no error behind. This paper describes the ingestion and ontology-tagging layer that turns a validated extraction stream into a knowledge graph of 537,157 entities and 2,198,567 relationships drawn from 98,795 government documents. We describe a record-identity ladder that decides sameness from identifier columns, name columns, display names and type-scoped position rather than from name similarity. The ladder governs de-duplication within parsed tables, while the graph write applies a coarser canonical-name key, so records sharing a canonical name merge automatically on exact equality. We argue rather than demonstrate that this is where the automation line belongs: no identity benchmark is reported, and the over-merges the key permits are undetectable by construction. That policy, under which entity resolution only ever flags candidates, followed an incident in which two surface forms of one name were merged, corrupting a correct record and deleting eight entities from an unrelated document. We then describe multi-class ontology tagging and an evidence asymmetry we did not anticipate: an entity name is an instance label rather than a type assertion, so matching name fragments against a class index invents classifications. Requiring anchored evidence cut role assignments on an enriched sample from 36 to 4, all confirmed correct. We quantify the graph's conformance debt, show secondary classifications compensating for a mis-parented primary class, and describe a curation queue grown to 48,403 pending proposals against 775 human decisions.
Vaibhav Dangaich, Kevin Lewis, Kundeshwar Pundalik· 0 citations
Constructing formal ontologies from domain documents requires simultaneously enforcing corpus grounding, vocabulary consistency, axiom-level expressivity, and end-to-end provenance, a combination no existing automatic system delivers. We present a seven-stage graph-grounded pipeline that converts domain documents into a complete, auditable Web Ontology Language (OWL) Terminological Box (TBox) without any unconstrained generation step. Documents are first encoded as Unified Discourse-Hypergraphs (UDH) capturing entity participation and discourse dependencies; subsequent stages transform this graph evidence into a class hierarchy, typed object and datatype properties, and restriction axioms, with Large Language Model (LLM) usage restricted to narrow, graph-grounded mediation tasks. A paired Assertional Box (ABox) population procedure grounds named individuals in the induced TBox, enabling SPARQL-based functional evaluation. Every emitted term carries a full decision chain from raw source passages through each pipeline stage, making the TBox directly auditable and suitable for targeted human refinement. Evaluated on the life insurance domain using two established benchmarks and a new 100-contract corpus spanning ten product types, our pipeline achieves strong results across all evaluation dimensions, outperforming direct and multi-agent LLM baselines on competency-question (CQ) coverage (0.85 vs. 0.63 and 0.62 on one term-life contract; 0.77 vs. 0.40 and 0.44 on another contract), while also attaining high keyphrase coverage comparable to a manually-constructed reference ontology and strong performance on structured gap-and-overlap reasoning, all without any manual TBox engineering. Ontology growth analysis provides evidence consistent with vocabulary saturation at scale, demonstrating that the pipeline produces stable, reusable domain representations from large document corpora.
Maruf Ahmed Mridul, A. Talukder, O. Seneviratne· 0 citations
A two-stage LLM-assisted workflow for French maintenance regulations is presented: ontology engineering from a SEMLEG-based core ontology, followed by construction of an ontology-grounded French legal knowledge graph.
Génesis Montenegro, M. Billami, Catherine Faron et al.· 0 citations
Integrating newly emerging terms into existing ontologies is a recurring maintenance problem in knowledge engineering, particularly in biomedical domains where terminology evolves faster than manual curation can accommodate. This paper addresses the placement of concepts that are absent from the target ontology—the out-of-knowledge-base setting—in which a textual mention must be assigned one or more insertion positions in the subsumption hierarchy rather than linked to an existing node. We propose a three-stage framework that extends the conventional retrieve-then-select paradigm with an intermediate stage of edge generation and enrichment, which expands the candidate set by traversing the local structure of the ontology. Stage 1 retrieves candidate edges using a fine-tuned bi-encoder trained with a max-margin objective; Stage 2 constructs and structurally enriches candidate edges; Stage 3 selects among them using either a fine-tuned cross-encoder or a large language model under explainable instruction tuning. We evaluate on two datasets derived from SNOMED CT, MM-S14-Disease and MM-S14-CPP, under a strict out-of-knowledge-base protocol. Fine-tuned pre-trained language models outperform zero-shot and instruction-tuned large language models on ranking accuracy, while the instruction-tuned configuration produces expert-auditable justifications at a modest cost in accuracy. On MM-S14-Disease, the strongest configuration places a correct insertion edge among the ten highest-ranked candidates for 38.7% of test mentions and recovers the complete gold edge set for 16.4%, against 26.1% and 9.2% for retrieval alone. The framework is positioned as decision support for ontology curators rather than as an autonomous ontology generator.
Zhanna B. Sadirmekova, M. Sambetbayeva, B. Abdygalym et al.· Information· 0 citations
A novel, interpretable metric for intrinsic KG quality assessment that measures how closely an automatically extracted graph approximates an"ideal"graph capturing the key noun phrases, predicate relations, and basic linguistic phenomena such as negation expressed in the source text is proposed.
Nipun Misra, Vikranth Udandarao, Aanchal Gupta et al.· 0 citations