[ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/assertion", "@graph": [ { "@id": "https://doi.org/10.48550/arXiv.2503.16131", "http://purl.org/dc/terms/title": [ { "@value": "MKG-Rank: Enhancing Large Language Models with Knowledge Graph for Multilingual Medical Question Answering" } ], "http://purl.org/spar/cito/describes": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#CachingMechanism" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#DeclarativeConversion" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#MKGRank" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#MultiAngleRanking" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#SelfInformationMining" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#WordLevelTranslationMechanism" } ], "http://purl.org/spar/cito/discusses": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#BM25" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#MedCPTCrossEncoder" }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#UMLSBertEmbeddings" } ], "@type": [ "http://www.w3.org/ns/prov#Entity" ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#CachingMechanism", "http://purl.org/dc/terms/subject": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/categories#KGEnhancedLLMInference" } ], "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#comment": [ { "@value": "This mechanism optimizes the retrieval of external medical KGs by storing previously retrieved KGs in a local knowledge base. It significantly reduces retrieval times during LLM inference, directly contributing to the efficiency of the overall KG-enhanced LLM system." } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "Caching Mechanism" } ], "https://neverblink.eu/ontologies/llm-kg/hasTopCategory": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/top-categories#KGEnhancedLLM" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#DeclarativeConversion", "http://purl.org/dc/terms/subject": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/categories#KGEnhancedLLMInference" } ], "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#comment": [ { "@value": "This mechanism converts raw medical KG triplets into LLM-digestible declarative sentences. It allows the LLM to better integrate and reason with external knowledge during inference, improving its ability to utilize the provided medical facts." } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "Declarative Conversion" } ], "https://neverblink.eu/ontologies/llm-kg/hasTopCategory": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/top-categories#KGEnhancedLLM" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#MKGRank", "http://purl.org/dc/terms/subject": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/categories#KGEnhancedLLMInference" } ], "http://purl.org/spar/fabio/hasURL": [ { "@value": "https://anonymous.4open.science/r/MKG-Rank-6B72", "@type": "http://www.w3.org/2001/XMLSchema#anyURI" } ], "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#comment": [ { "@value": "This framework enhances English-centric LLMs to perform multilingual medical QA by integrating comprehensive external medical knowledge graphs during the inference stage. It bridges language gaps and provides relevant, up-to-date knowledge to LLMs for better reasoning." } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "MKG-Rank" } ], "https://neverblink.eu/ontologies/llm-kg/hasTopCategory": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/top-categories#KGEnhancedLLM" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#MultiAngleRanking", "http://purl.org/dc/terms/subject": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/categories#KGEnhancedLLMInference" } ], "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#comment": [ { "@value": "This strategy selects the most relevant medical triplets from retrieved KGs by ranking them based on similarity with the question (using UMLS-BERT embeddings) and further filtering with a MedCPT Cross Encoder. It ensures the LLM receives high-quality, pertinent information for accurate inference." } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "Multi-Angle Ranking" } ], "https://neverblink.eu/ontologies/llm-kg/hasTopCategory": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/top-categories#KGEnhancedLLM" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#SelfInformationMining", "http://purl.org/dc/terms/subject": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/categories#KGEnhancedLLMInference" } ], "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#comment": [ { "@value": "When external KG retrieval is ineffective, this method employs the BM25 algorithm to extract relevant world knowledge from the LLM's own internal representations. It acts as a fallback strategy to ensure the LLM consistently has background information for medical QA, enhancing its robustness during inference." } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "Self-Information Mining" } ], "https://neverblink.eu/ontologies/llm-kg/hasTopCategory": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/top-categories#KGEnhancedLLM" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#WordLevelTranslationMechanism", "http://purl.org/dc/terms/subject": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/categories#KGEnhancedLLMInference" } ], "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#comment": [ { "@value": "This mechanism, part of MKG-Rank, extracts medical entities from multilingual questions and options using an LLM, then translates them into English. This enables the LLM to leverage English-centric KGs for multilingual QA, enhancing its performance during inference." } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "Word-level Translation Mechanism" } ], "https://neverblink.eu/ontologies/llm-kg/hasTopCategory": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/top-categories#KGEnhancedLLM" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#BM25", "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "BM25" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#MedCPTCrossEncoder", "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "MedCPT Cross Encoder" } ] }, { "@id": "https://neverblink.eu/ontologies/llm-kg/methods#UMLSBertEmbeddings", "@type": [ "http://purl.org/spar/fabio/Workflow" ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "UMLS-BERT embeddings" } ] } ] }, { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/pubinfo", "@graph": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0", "http://purl.org/dc/terms/created": [ { "@value": "2026-02-26T15:55:27.464Z", "@type": "http://www.w3.org/2001/XMLSchema#dateTime" } ], "http://purl.org/dc/terms/creator": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/agent" } ], "http://purl.org/nanopub/x/hasNanopubType": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/PaperAssessmentResult" } ], "http://www.w3.org/2000/01/rdf-schema#label": [ { "@value": "LLM-KG assessment for paper 10.48550/arXiv.2503.16131" } ] }, { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/sig", "http://purl.org/nanopub/x/hasAlgorithm": [ { "@value": "RSA" } ], "http://purl.org/nanopub/x/hasPublicKey": [ { "@value": "MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAwNz2QK3SEifno78S7+48zUB0xpTex3mAzW73ZimHqNcdEMU5/apslrGrTHGFAt/Chocgo++r6JQp5ygY7NyJHGWdaIqnt85pjX4PbNfLAvapyUO00qZP34fY61w4eZ9UMtleWEsmZKRtQPyJ8ODl46i/rfPuZlcJGpM9Nmy5mpGWuepqIEvF4a/t7pLVeCEDFSYXT+yaiygt6ynIK5f7TtEDhZpeUf/Q74WhMPJXm4yTU/hqOX4IW+50kWHNArGGZwUaXwzyG6M3Zd6UMModryGkLqS4H/MSE3ZA1Ylnms7BfWLEXhMWlaKi6HRV4nGRDLhxVSi9LSRi3LWKLhNIIQIDAQAB" } ], "http://purl.org/nanopub/x/hasSignature": [ { "@value": "N2UbPmY5xFtgVOnItcOMYoJbF1sO6k2Ix2/gm5wQngPckyNYrDlQPa5oii1fbKkm3cToZsokZDhdYD7t8Ut/TFP2VR8PNODfRBaVkrKb3N+WrJAwHCLyA3DzkBxww7zqv2B0eYlwJjmwdSNAFR+DyBvpj9SboJL58MX4b0nm//u4g+EwPehDRmlaXafGrMZnML5CUY0U1FkmBe4LlqSHBM+NtooVpnS2/nz1KA4BK7GxrB7xejXB2MZOtfh/OMh3B1b9aJ64uB1wdlVuJhPEmGabeXFepE6RxUo0huyMuwHR8NtEjogX+ZkD4qiYCycN/KbOlYi/rESa6e94J1O3fw==" } ], "http://purl.org/nanopub/x/hasSignatureTarget": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0" } ], "http://purl.org/nanopub/x/signedBy": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/agent" } ] } ] }, { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/Head", "@graph": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0", "http://www.nanopub.org/nschema#hasAssertion": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/assertion" } ], "http://www.nanopub.org/nschema#hasProvenance": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/provenance" } ], "http://www.nanopub.org/nschema#hasPublicationInfo": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/pubinfo" } ], "@type": [ "http://www.nanopub.org/nschema#Nanopublication" ] } ] }, { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/provenance", "@graph": [ { "@id": "https://w3id.org/np/RA9UhFFrfgd2Uu4NfZsny8CxPYI-w5YdYgs8B41DKAIs0/assertion", "http://www.w3.org/ns/prov#wasAttributedTo": [ { "@id": "https://neverblink.eu/ontologies/llm-kg/agent" } ], "http://www.w3.org/ns/prov#wasDerivedFrom": [ { "@id": "https://doi.org/10.48550/arXiv.2503.16131" } ] } ] } ]