关键词: 关系数据库, 信息检索, 对象级别检索

Abstract: Object-level information retrieval has been a hot research topic recently. To solve the problem, this paper proposes an object-level information retrieval method for relational databases, named as DBORank, which can improve the effect of information retrieval over relational databases. DBORank designs a flexible and effective scoring mechanism from both databases and information retrieval perspectives, which takes into account not only the link structure of object-level data graph, but also the internal structure of object nodes, the type and weight of edges, the relevance of object content, etc. At the same time, it optimizes the iteration algorithm of object scoring. The experimental results show that DBORank performs efficiently.