Original scientific paper

https://doi.org/10.17559/TV-20160419053402

Keyword Search in Large-Scale Databases with Topic Cluster Units

Yingqi Wang orcid.org/0000-0002-4215-5903 ; School of Computer Science and Technology, Harbin Engineering University Harbin, 150001, China
Nianbin Wang ; School of Computer Science and Technology, Harbin Engineering University Harbin, 150001, China
Lianke Zhou orcid.org/0000-0002-3601-9286 ; School of Computer Science and Technology, Harbin Engineering University Harbin, 150001, China

Abstract

To solve the inefficiency of the existing keyword search methods in large databases, this paper proposes TCU-based query, an offline query method based on topic cluster units. First, topic cluster units (TCUs) are constructed through vertical grouping and horizontal grouping on tables and tuples. In contrast to traditional keyword query methods, this offline method cannot only reduce the query response time, but also return results comprising richer and more complete semantic information. In order to further improve the efficiency of data preprocessing, an optimized solution for table join ordering based on the genetic algorithm is presented. Second, we select index terms using the association rule, and then we build an index on every topic cluster; by doing so we can improve the query speed significantly. Finally, we conduct extensive experiments to demonstrate that our approach greatly improves the performance of keyword search.

Keywords

clustering; keyword search; relational databases; subject indexes; topic cluster units

Hrčak ID:

202609

URI

https://hrcak.srce.hr/202609

Publication date:

28.6.2018.

Visits: 1.669 *