2020년 컴퓨터종합학술대회
Current Result Document :
| 한글제목(Korean Title) |
네임드 그래프의 효율적인 빈발 패턴 관리 및 압축 시스템 |
| 영문제목(English Title) |
Efficient Frequent Pattern Management and Compression System in Multiple Named Graphs |
| 저자(Author) |
탄지나 술타나
구두스 우매르
무함마드 우매르
김태연
골람 모르셰드
이영구
Tangina Sultana
Umair Qudus
Muhammad Umair
Taeyeon Kim
Md Golam Morshed
Young-Koo Lee
|
| 원문수록처(Citation) |
VOL 47 NO. 01 PP. 0038 ~ 0040 (2020. 07) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
The increasing volume of RDF data demands efficient compressed data structure to compactly represent RDF triples and fast pattern matching operations. Header-Dictionary-Triple (HDT) is capable of compressing large datasets without prior decompression but it cannot handle the data redundancy and multiple RDF named graphs. In this study, we propose a novel approach to extend the HDT by introducing compressed hash table. It can reduce the data redundancy by identifying the frequent graph pattern in the dataset and detect predicates in the RDF graphs. Moreover, we use graph interpreter to detect the data in named graphs more easily. After evaluation on the real world datasets, simulation results affirm that our proposed approach substantially compresses the RDF graphs and performs better than the HDT. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
