TY - GEN
T1 - Keyword-Centric Community Search over Large Heterogeneous Information Networks
AU - Qiao, Lianpeng
AU - Zhang, Zhiwei
AU - Yuan, Ye
AU - Chen, Chen
AU - Wang, Guoren
N1 - Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - Community search in heterogeneous information networks (HINs) has attracted much attention in recent years and has been widely used for graph analysis works. However, existing community search studies over heterogeneous information networks ignore the importance of keywords and cannot be directly applied to the keyword-centric community search problem. To deal with these problems, we propose kKP -core, which is defined based on a densely-connected subgraph with respect to the given keywords set. A kKP -core is a maximal set of P -connected vertices in which every vertex has at least one KP -neighbor and k path instances. We further propose three algorithms to solve the keyword-centric community search problem based on kKP -core. When searching for answers, the basic algorithm Basic- kKP -core will enumerate all paths rather than only the path instances of the given meta-path P. To improve efficiency, we design an advanced algorithm AdvkKP -core using a new method of traversing the search space based on trees to accelerate the searching procedure. For online queries, we optimize the approach with a new index to handle the online queries of community search over HINs. Extensive experiments on HINs are conducted to evaluate both the effectiveness and efficiency of our proposed methods.
AB - Community search in heterogeneous information networks (HINs) has attracted much attention in recent years and has been widely used for graph analysis works. However, existing community search studies over heterogeneous information networks ignore the importance of keywords and cannot be directly applied to the keyword-centric community search problem. To deal with these problems, we propose kKP -core, which is defined based on a densely-connected subgraph with respect to the given keywords set. A kKP -core is a maximal set of P -connected vertices in which every vertex has at least one KP -neighbor and k path instances. We further propose three algorithms to solve the keyword-centric community search problem based on kKP -core. When searching for answers, the basic algorithm Basic- kKP -core will enumerate all paths rather than only the path instances of the given meta-path P. To improve efficiency, we design an advanced algorithm AdvkKP -core using a new method of traversing the search space based on trees to accelerate the searching procedure. For online queries, we optimize the approach with a new index to handle the online queries of community search over HINs. Extensive experiments on HINs are conducted to evaluate both the effectiveness and efficiency of our proposed methods.
KW - Community
KW - Heterogeneous information networks
KW - Keyword-centric
UR - http://www.scopus.com/inward/record.url?scp=85104740410&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-73194-6_12
DO - 10.1007/978-3-030-73194-6_12
M3 - Conference contribution
AN - SCOPUS:85104740410
SN - 9783030731939
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 158
EP - 173
BT - Database Systems for Advanced Applications - 26th International Conference, DASFAA 2021, Proceedings
A2 - Jensen, Christian S.
A2 - Lim, Ee-Peng
A2 - Yang, De-Nian
A2 - Lee, Wang-Chien
A2 - Tseng, Vincent S.
A2 - Kalogeraki, Vana
A2 - Huang, Jen-Wei
A2 - Shen, Chih-Ya
PB - Springer Science and Business Media Deutschland GmbH
T2 - 26th International Conference on Database Systems for Advanced Applications, DASFAA 2021
Y2 - 11 April 2021 through 14 April 2021
ER -