GraphSUM: scalable graph summarization for efficient question answering

Nasrin Shabani, Amin Beheshti, Jia Wu, Maryam Khanian Najafabadi, Jin Foo, Alireza Jolfaei

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contribution

5 Downloads (Pure)

Abstract

Efficiently processing large-scale graphs for question-answering tasks presents a significant challenge, given the complexity and volume of data involved in such graphs. This paper presents a new framework that combines attention-based graph summarization with innovative graph sampling methods designed specifically for large-scale graph processing and question-answering applications. Our approach excels in its ability to process large-scale graphs efficiently, leveraging effective sampling and attention mechanisms to enhance feature extraction. A key aspect of our approach is graph summarization techniques, which concentrate on essential information, boosting the accuracy and computational efficiency of question answering. Our framework proves its efficacy in real-world scenarios through practical demonstrations, notably within academic databases. This showcases a substantial advancement in information retrieval and graph-based data navigation, marking a significant leap forward in the field.

Original languageEnglish
Title of host publicationProceedings of the 27th International Conference on Extending Database Technology (EDBT), 25th March-28th March, 2024
Place of PublicationOnline
PublisherOpenProceedings.org
Pages794-797
Number of pages4
ISBN (Electronic)9783893180950
DOIs
Publication statusPublished - 18 Mar 2024
Event27th International Conference on Extending Database Technology, EDBT 2024 - Paestum, Italy
Duration: 25 Mar 202428 Mar 2024

Publication series

NameAdvances in Database Technology - EDBT
Number3
Volume27
ISSN (Electronic)2367-2005

Conference

Conference27th International Conference on Extending Database Technology, EDBT 2024
Country/TerritoryItaly
CityPaestum
Period25/03/2428/03/24

Bibliographical note

Copyright the Author(s) 2024. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Fingerprint

Dive into the research topics of 'GraphSUM: scalable graph summarization for efficient question answering'. Together they form a unique fingerprint.

Cite this