面向top-K分级的数据库关键词查询系统体系结构

Top-K-oriented hierarchical keyword-based information query system architecture over databases

关系数据库关键词查询系统目前存在的主要问题是查询效率不高、查询效果不令人满意,其根源在于查询系统的体系结构不合理。提出了一种新型的面向top-分级的数据库关键词查询系统体系结构,并用提出的新的评分函数公式、基于索引的候选网络预处理方法、绑定实时过滤阈值的top-搜索算法对该体系结构进行实例化。结果显示该系统结构具有很好的可行性,并且由实例化后的体系结构实现的系统具有很好的查询性能。

It is the irrational system architecture that leads to the problems of inefficient retrieval and unsatisfied query results （in keyword query system over relational database）. In order to solve the above two problems, in this paper, a top-k-oriented hierarchical keyword-based information retrieval system architecture over databases is initiated. Such an architecture is instantiated by the evaluation formula, preparation technique of index-based candidate network and top-k query algorithm of binding real-time threshold illtration. The results verify the great feasibility of this architecture. Furthermore, the query system based on this architecture can achieve good query performance.