Exploring In-Context Learning Capabilities of Foundation Models for Generating Knowledge Graphs from Text

Simple item page
AudienciaPúblico en generales_ES
CoberturaMéxico[ 143]es_ES
Fecha de ingreso2026-10-05T16:34:59Z
Fecha de publicación2023-01-01
ResumenKnowledge graphs can represent information about the real-world using entities and their relations in a structured and semantically rich manner and they enable a variety of downstream applications such as question-answering, recommendation systems, semantic search, and advanced analytics. However, at the moment, building a knowledge graph involves a lot of manual effort and thus hinders their application in some situations and the automation of this process might benefit especially for small organizations. Automatically generating structured knowledge graphs from a large volume of natural language is still a challenging task and the research on sub-tasks such as named entity extraction, relation extraction, entity and relation linking, and knowledge graph construction aims to improve the state of the art of automatic construction and completion of knowledge graphs from text. The recent advancement of foundation models with billions of parameters trained in a self-supervised manner with large volumes of training data that can be adapted to a variety of downstream tasks has helped to demonstrate high performance on a large range of Natural Language Processing (NLP) tasks. In this context, one emerging paradigm is in-context learning where a language model is used as it is with a prompt that provides instructions and some examples to perform a task without changing the parameters of the model using traditional approaches such as fine-tuning. This way, no computing resources are needed for re-training/fine-tuning the models and the engineering effort is minimal. Thus, it would be beneficial to utilize such capabilities for generating knowledge graphs from text. In this paper, grounded by several research questions, we explore the capabilities of foundation models such as ChatGPT to generate knowledge graphs from the knowledge it captured during pre-training as well as the new text provided to it in the prompt. The paper provides a qualitative analysis of a set of example outputs generated by a foundation model with the aim of knowledge graph construction and completion. The results demonstrate promising capabilities. Furthermore, we discuss the challenges and next steps for this research work.es_ES
URIhttps://riuat.uat.edu.mx/handle/123456789/6460
IdiomaInglés[20]es_ES
EditorialCEUR-WSes_ES
RelaciónCEUR Workshop Proceedingses_ES
URL relacionadohttps://www.scopus.com/pages/publications/85168818600?origin=resultslistes_ES
DerechosAcceso restringido / Suscripción (Metadatos de producción científica)es_ES
Licenciahttp://purl.org/coar/access_right/c_16eces_ES
FuenteCEUR Workshop Proceedings
Palabra claveFoundation Modelses_ES
Palabra claveIn-Context Learninges_ES
Palabra claveKnowledge Graph Completiones_ES
Palabra claveKnowledge Graph Constructiones_ES
Palabra claveLarge Language Modelses_ES
Palabra claveOntologyes_ES
Palabra claveRelation Extractiones_ES
TítuloExploring In-Context Learning Capabilities of Foundation Models for Generating Knowledge Graphs from Textes_ES
TipoPonenciaes_ES
ArbitradoHa sido Arbitradoes_ES
AutorKhorashadizadeh, Hanieh
AutorMihindukulasooriya, Nandana
AutorTiwari, Sanju
AutorGroppe, Jinghua
AutorGroppe, Sven
AutorKhorashadizadeh, Haniehes_ES
AutorMihindukulasooriya, Nandanaes_ES
AutorTiwari, Sanjues_ES
AutorGroppe, Jinghuaes_ES
AutorGroppe, Svenes_ES
InstituciónUniversidad Autónoma de Tamaulipas
InstituciónUniversidad Autónoma de Tamaulipases_ES
Rango de páginas132-153es_ES
URL relacionadahttps://www.scopus.com/pages/publications/85168818600?origin=resultslist
Tipo de artículoIndexado
Tipo de artículoIndexadoes_ES
Volumen3447es_ES

Files

Collections

Documentación Cientifica