A Deep Neural Approach to KGQA via SPARQL Silhouette Generation

Purkayastha, Sukannya; Dana, Saswati; Garg, Dinesh; Khandelwal, Dinesh; Bhargav, G P Shrivatsa

doi:10.1109/ijcnn55064.2022.9892263

Cited by 9 publications

(1 citation statement)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent work on KGs has generated KGs in different domains whereas some works perform querying over the existing KGs to improve the tasks like recommendation systems, link prediction, node classification, and knowledge discovery [34][35][36]. Another line of works in [37][38][39][40] generate SPARQL queries from natural language for existing KGs like DBpedia, Wikidata for complex querying. Furthermore, Graph Neural Network-based learning models have been employed to enhance KGs' performance in tasks such as link prediction and multihop querying [41][42][43].…”

Section: Related Work and Motivationmentioning

confidence: 99%

Q-KGSWS: Querying the Hybrid Framework of Knowledge Graph and Semantic Web Services for Service Discovery

Thapar,

Sharma

2024

IJACSA

View full text Add to dashboard Cite

In the era of big data, Knowledge Graphs (KGs) have become essential tools for managing interconnected datasets across various domains. This paper introduces a novel RDF (Resource Description Framework) based Knowledge Graph of Semantic Web Services (KGSWS), designed to enhance service discovery. Leveraging the versatile SPARQL query language, the framework facilitates precise querying operations on KGSWS, enabling customized service matching for user queries. Through comprehensive experimentation and analysis, notable improvements in accuracy (69.75% and 90.01%) and rapid response times (0.61s and 1.57s) across two semantic search levels are demonstrated, validating the efficacy of the approach. Furthermore, research questions regarding the interlinking of ontologies, methods for formulating automatic queries, and efficient retrieval of services are addressed, offering insights into future avenues for research. This work represents a significant advancement in the domain of semantic web services, with potential applications across various industries reliant on efficient service identification and integration. Future phases of research will focus on logical inference and the integration of machine learning-based graph embedding models, promising even greater strides in knowledge discovery within the KGSWS framework, thus reshaping the domain of semantic web services.

show abstract