中文摘要：

为了实现信息网数据库管理系统INMDBMS（Information Network Database Management System）的逻辑推理功能,提出了适用于INMDBMS的信息规则语言IRL（Information Rule Language）用于规则的表示。总结Datalog语言中几种传统的递归查询算法,同时结合INMDBMS的数据模型特点,设计并实现了IGQA（INMDB Goal_driven Recursive Query Algorithm）作为IRL规则语言的递归查询算法。IGQA以深度优先、回退的方式来实现递归查询。在真实数据集上的实验表明,当数据规模增长时,IGQA具有较高的执行成功率和较好的执行时间稳定性。

英文摘要：

In order to implement the logic reasoning capability of information network database management system （ INMDBMS）, we proposed the information rule language （IRL）, which is applicable to INMDBMS, for representing the rules. We summarised several traditional recursive query algorithms in Datalog language and combined at the same time with the characteristics of data model of INMDBMS, designed and implemented the INMDB goal_driven recursive query algorithm （IGQA） as the recursive query algorithm of IRL. IGQA achieves the recursive query in the way of depth-first and rollback. Experiments on real datasets showed that when the size of data grew, IGQA had a higher success rate and better time stability in execution.