【24h】

THE RULES ENGINE DESIGN AND IMPLEMENTATION BASED ON THE IMPROVED RETE ALGORITHM

机译:基于改进后的算法的规则引擎设计与实现

获取原文

摘要

Current rules engine is mainly designed based on the Rete algorithm. Rete algorithm is the most efficient forward inference algorithm, but it has some shortcomings :the cost of deleting WME is the same with adding it;and β network has too many redundant nodes to save the partial match results which waste a lot of space; what’s more null join often occur. Therefore, this paper will improve the shortcomings of the Rete algorithm,put forward a timestamp-βmemory upper bound method to restrict the β network space and implement the asymmetric deleting WME, and use the unjoin method to solve the null join problem. We design the rules engine based on the improved algorithm. Then give the empirical result to show its feasibility and superiority. At last,we give the future work we should do to improve ulteriorly.
机译:当前规则引擎主要是基于Rete算法设计的。 Rete算法是最有效的前向推理算法,但有一些缺点:删除WME的代价与添加WME的代价相同;β网络的冗余节点过多,无法保存部分匹配结果,从而浪费大量空间。更常见的是空连接。因此,本文将弥补Rete算法的不足,提出一种时间戳-β内存上限方法来限制β网络空间,实现非对称删除WME,并使用unjoin方法解决空联接问题。我们基于改进算法设计规则引擎。然后给出实证结果,说明其可行性和优越性。最后,我们给出了今后应做的工作,以进一步改善。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号