首页> 外文会议>International conference on logic for programming, artificial intelligence, and reasoning >Incremental Tabling for Query-Driven Propagation of Logic Program Updates
【24h】

Incremental Tabling for Query-Driven Propagation of Logic Program Updates

机译:用于逻辑驱动程序更新的查询驱动传播的增量制表

获取原文

摘要

We foster a novel implementation technique for logic program updates, which exploits incremental tabling in logic programming - using XSB Prolog to that effect. Propagation of updates of fluents is controlled by initially keeping any fluent updates pending in the database. And, on the initiative of queries, making active just those updates up to the timestamp of an actual query, by performing incremental assertions of the pending ones. These assertions, in turn, automatically trigger system-implemented incremental bottom-up tabling of other fluents (or their negated complements), with respect to a predefined overall upper time limit, in order to avoid runaway iteration. The frame problem can then be dealt with by inspecting a table for the latest time a fluent is known to be assuredly true, i.e., the latest time it is not supervened by its negated complement, relative to the given q'uery time. To do so, we adopt the dual program transformation for defining and helping propagate, also incrementally and bottom-up, the negated complement of a fluent, in order to establish whether a fluent is still true at some time point, or rather if its complement is. The use of incremental tabling in this approach affords us a form of controlled, but automatic, system level truth-maintenance, up to some actual query time. Consequently, propagation of update side-effects need not employ top-down recursion or bottom-up iteration through a logically defined frame axiom, but can be dealt with by the mechanics of the underlying world. Our approach thus reconciles high-level top-down deliberative reasoning about a query, with autonomous low-level bottom-up world reactivity to ongoing updates, and it might be adopted elsewhere for reasoning in logic.
机译:我们培育了一种用于逻辑程序更新的新颖实现技术,该技术利用逻辑编程中的增量制表法-为此使用XSB Prolog。通过最初在数据库中保留任何流利更新来控制流利更新的传播。并且,在查询的主动下,通过执行未决查询的增量断言,仅在实际查询的时间戳之前激活那些更新。这些断言又反过来相对于预定义的总体上限自动触发其他流利系统(或它们的否定补码)的系统实现的自下而上的制表,从而避免失控的迭代。然后可以通过检查一张桌子来解决框架问题,最晚的时间可以确定是流利的,该时间肯定是正确的,也就是说,相对于给定的查询时间,它不能被其否定的补码所取代。为此,我们采用双重程序转换来定义和帮助对流利的否定补语进行定义和帮助(增量地和自下而上地)传播,以便确定流利性在某个时间点是否仍然正确,或者更确切地说是否是其补语是。在这种方法中使用增量制表为我们提供了一种受控的,但自动的系统级真相维护形式,可以保留一些实际查询时间。因此,更新副作用的传播不需要通过逻辑定义的框架公理采用自上而下的递归或自下而上的迭代,而是可以由基础世界的机制来处理。因此,我们的方法使查询的高级自上而下的协商推理与正在进行的更新的自治的低级自下而上的世界反应性协调一致,并且可能在其他地方被采用以进行逻辑推理。

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