缺省推理与认识进程

本文概述了一个可以刻画知识的增长、更新以及假说的进化的开放逻辑理论;给出了有关新假设、事实反驳、假说的重构、认识进程及其极限等概念,讨论了它们的性质并证明了与之有关的定理。本文对开放逻辑和Reiter缺省推理理论做了比较研究,并用开放逻辑的概念给出了缺省的一个模型论解释,给出了扩充的构造,并证明了Reiter缺省证明概念的完全性。     

开放逻辑—一个刻画知识增长和更新的逻辑理论

本文建立了一个可以刻画知识的增长、更新以及假说的进化的逻辑理论;给出了新假设、事实反驳、假说的重构及认识进程等概念的定义,讨论了它们的性质并证明了与之有关的定理,本文还定义了认识进程的极限概念并证明任何关于某一特定问题的经验模型都是一认识进程的极限。作为开放逻辑的应用,本文给出了 Reiter 缺省推理理论的一个模型论解释。     

拓展溯因逻辑程序设计的辩论理论基础

讨论一类扩展的溯因程序,它包含经典否定、缺省否定、一致性约束以及溯因推理机制,论文的主要思想是：⑴本文限制某些假说（包括不相容假说）的攻击能力这样的出发点,定义一种更符合直观理解的“反驳”与“击败”概念,其中的Ⅱ－型击败关系具有动态的特点;⑵首次尝试一致性约束可以引起假说间的反驳与是一致笥约束不再仅仅是全局性的相容约束,基于这些思想提出的完全类语义拓展了Ｄｕｎｇ所奠定的溯因逻辑程序设计的辩论理论基     

神经网络问题求解机制

本文提出了神经网络知识表示的形式化描述语言和知识单元的概念，用于给出传统符号逻辑中的概念，用于给出传统符号逻辑中的概念，属性及它们之间的层次关系如何在神经网络中进行表示，提出了激活强度的度量标准，从而在理论上给出了ＮＮ中继承和识别问题的形式化处理方法。在此基础上，提出了ＮＮ正向和反向问题求解机制。这里，值得一提的是，本文通过在某一特定领域中对ＮＮ推理机制的成功探讨，有力地反驳了人们对于ＮＮ模型的一     

开放的限制理论

本文基于开放逻辑理论，建立了一阶谓词限制理论的知识增长、更新及理论进化的开放的限制理论，给出了谓词限制理论中新假设、事实反驳及伪事实反驳、Ｃ－重构、Ｃ－认识进程及其极限的定义，讨论了它们的性质并证明了有关的定理，进而描述了限制理论的动态特征－Ｃ－认识进程，证明了其极限定量，并比较了它与一般认识进程及限制理论的关系。     

开放的缺省理论

本文基于开放逻辑理论，给出了缺省理论Ｔ＝＜Ｄ，Ｗ＞扩充Ｅ的新假设，事实反驳、ｅ－重构、ｅ－认识进程及其极限等概念的意义，讨论了Ｗ变化时新扩充的变化规律，并证明了相关的定理，本文还建立了缺省理论的一个动态描述过程，证明了其极限是某一特定问题的经验公式集，最后与相关工作进行了比较。     

扩充逻辑程序设计的R－演算──知识库维护的操作方法

本文首先介绍了知识库维护过程中诸如知识库序列、新规则、用户反驳以及重构等概念；然后给出了一个扩充逻辑程序设计的框架，在这一框架下，每个逻辑程序等价于一个知识库；进一步定义了一个转换系统，称为扩充逻辑程序设计的Ｒ－演算，对一个给定的知识库和用户反驳，此演算可以导出知识库的最佳修正；同时证明了该演算的可靠性和完备性；另外，对本文的工作与其他相关工作进行了比较；最后，给出了本文的结论．     

用扩充的面向对象模型表达抽象概念的方法

给出了抽象函数类的面向对象表示方法，用引入自由变量的概念扩充面向对象的建模方法，同时讨论了面向对象的模型中自由变量的性质，并介绍了模糊对象建模中有关抽象函数类的构造方法。     

布尔算子模糊逻辑中一种新的假说修正方法

在基于具有某种程度的不一致性的知识进行推理的过程中，为了消除知识的不一致性，只能将这些知识看成假说，并通过对假说进行修正以重新获得一个一致的假说。文中基于布尔算子模糊逻辑，给出了一种新的假说修正方法，这种方法能够在一定程度上更多地保留被修正知识的合理成分，以便在以后有机会得到恢复。     

Web模糊聚类方法及其应用

本文提出了Web模糊聚类的概念，给出了Web模糊聚美的过程模型WFCM并进行了详细阐述，沦述了Web模糊聚类在Web访问信息挖掘中，尤其是在Web用户聚类和Web页面聚类方面的应用．最后用实例证明了在Web页面聚类中使用Web模糊聚类的可行性。     

JBInsTrace: A tracer of Java and JRE classes at basic-block granularity by dynamically instrumenting bytecode

Understanding what happens during the runtime of a Java program is difficult. Tracking runtime flow can bring valuable information for program understanding and behavior analysis. Polymorphism, thread concurrency or even simple facts like the number of method invocations and the number of executed bytecodes are valuable information to track, but are difficult to compute outside the Java Virtual Machine (JVM) on running programs. In this paper, we present JBInsTrace, a new tool that instruments and traces Java bytecode. It produces static information about source code and a very fine grained trace of Java software execution, combining them to allow detailed analysis of the runtime. Our tool differs from others because it does not only trace program classes but also JRE classes, and does so at basic block level, without altering the JVM and without statically modifying class files. We explain JBInsTrace design, focused towards efficiency, which results in reasonable performance penalty.     

The domain theory for requirements engineering

Retrieval, validation, and explanation tools are described for cooperative assistance during requirements engineering and are illustrated by a library system case study. Generic models of applications are reused as templates for modeling and critiquing requirements for new applications. The validation tools depend on a matching process which takes facts describing a new application and retrieves the appropriate generic model from the system library. The algorithms of the matcher, which implement a computational theory of analogical structure matching, are described. A theory of domain knowledge is proposed to define the semantics and composition of generic domain models in the context of requirements engineering. A modeling language and a library of models arranged in families of classes are described. The models represent the basic transaction processing or `use case' for a class of applications. Critical difference rules are given to distinguish between families and hierarchical levels. Related work and future directions of the domain theory are discussed     

电脑故障分类及其领域知识的表达

在电脑软硬件故障的分类和特点的基础上,抽象出其故障知识的表达模型。采用了基于模糊产生式规则的方法来表达其领域知识,同时阐述了该方法中模糊系数和权系数的确定过程,给出了一个典型的具体实例。证明了该方法的合理性和有效性,实现了一种与电脑故障实际情况相符的较为合理的知识表达方法。     

Mind change complexity of learning logic programs

The present paper motivates the study of mind change complexity for learning minimal models of length-bounded logic programs. It establishes ordinal mind change complexity bounds for learnability of these classes both from positive facts and from positive and negative facts. Building on Angluin's notion of finite thickness and Wright's work on finite elasticity, Shinohara defined the property of bounded finite thickness to give a sufficient condition for learnability of indexed families of computable languages from positive data. This paper shows that an effective version of Shinohara's notion of bounded finite thickness gives sufficient conditions for learnability with ordinal mind change bound, both in the context of learnability from positive data and for learnability from complete (both positive and negative) data. Let ω be a notation for the first limit ordinal. Then, it is shown that if a language defining framework yields a uniformly decidable family of languages and has effective bounded finite thickness, then for each natural number m>0, the class of languages defined by formal systems of length ⩽m:

• •is identifiable in the limit from positive data with a mind change bound of ω^m;
• •is identifiable in the limit from both positive and negative data with an ordinal mind change bound of ω×m.

The above sufficient conditions are employed to give an ordinal mind change bound for learnability of minimal models of various classes of length-bounded Prolog programs, including Shapiro's linear programs, Arimura and Shinohara's depth-bounded linearly covering programs, and Krishna Rao's depth-bounded linearly moded programs. It is also noted that the bound for learning from positive data is tight for the example classes considered.     

面向司法案件的案情知识图谱自动构建

以法学知识为中心的认知智能是当前司法人工智能发展的重要方向。该文提出了以自然语言处理(NLP)为核心技术的司法案件案情知识图谱自动构建技术。以预训练模型为基础,对涉及的实体识别和关系抽取这两个NLP基本任务进行了模型研究与设计。针对实体识别任务,对比研究了两种基于预训练的实体识别模型;针对关系抽取任务,该文提出融合平移嵌入的多任务联合的语义关系抽取模型,同时获得了结合上下文的案情知识表示学习。在“机动车交通事故责任纠纷”案由下,和基准模型相比,实体识别的F₁值可提升0.36,关系抽取的F₁值提升高达2.37。以此为基础,该文设计了司法案件的案情知识图谱自动构建流程,实现了对数十万份判决书案情知识图谱的自动构建,为类案精准推送等司法人工智能应用提供语义支撑。     

模糊粗糙集的相似度量和相似性方向

粗糙集理论是一种新的处理模糊和不确定性知识的软计算工具,在人工智能及认知科学等众多领域已经得到了广泛的应用。相似度量的研究是模糊集理论与粗糙集理论的热点问题之一。文章提出了一种更精确、更合理的相似度量方法,讨论了它的一些性质。然后,在此基础上提出了模糊粗糙集的相似性方向的概念,用于比较两个相似的模糊粗糙集所包含信息的精确性大小,并给出了一个关于相似性方向的判别函数。这在近似推理、模式识别和决策分析等领域有着广泛的应用。最后,通过一个实例,分析说明了这种相似度量方法和相似性方向的判别方法是更合理更有效的。     

重叠实体关系抽取综述

实体关系抽取能够从文本中提取事实知识,是自然语言处理领域中重要的任务。传统关系抽取更加关注于单实体对的关系,但是句子内包含不止一对实体且实体间存在重叠现象,因此重叠实体关系抽取任务具有重大研究价值。任务发展至今,总体可以分为基于序列到序列、基于图和基于预训练语言模型三种方式。基于序列到序列的方式主要以标注策略和复制机制的方法为主,基于图的方式主要以静态图和动态图的方法为主,基于预训练语言模型的方式主要以BERT挖掘潜在语义特征的方法为主。回顾该任务的发展历程,讨论分析每种模型的优势及不足点;结合目前研究的最近动态,对未来的研究方向进行展望。     

Class Based Contextual Logic for DOOD

Cntextual logic provides a mechanism to reason about modules.In this paper,this theory of modules if modules is extended to a context theory of classes where class is in the true spirit of object-oriented databases.The logic,referred to as CLOG,is class-based.CLOG supports class,object identity,multiple role of object, monotonic and non-monotonic inheritance of data and method,method factoring,views,derived and query classes.Views and derived classes are queries in themselves.Objects are pure data terms representing the ground instances of facts in the class.Object identity is a first class term in the logic.Inheritance is handled through delegation.     

Complexity theoretic hardness results for query learning

We investigate the complexity of learning for the well-studied model in which the learning algorithm may ask membership and equivalence queries. While complexity theoretic techniques have previously been used to prove hardness results in various learning models, these techniques typically are not strong enough to use when a learning algorithm may make membership queries. We develop a general technique for proving hardness results for learning with membership and equivalence queries (and for more general query models). We apply the technique to show that, assuming , no polynomial-time membership and (proper) equivalence query algorithms exist for exactly learning read-thrice DNF formulas, unions of halfspaces over the Boolean domain, or some other related classes. Our hardness results are representation dependent, and do not preclude the existence of representation independent algorithms.?The general technique introduces the representation problem for a class F of representations (e.g., formulas), which is naturally associated with the learning problem for F. This problem is related to the structural question of how to characterize functions representable by formulas in F, and is a generalization of standard complexity problems such as Satisfiability. While in general the representation problem is in , we present a theorem demonstrating that for "reasonable" classes F, the existence of a polynomial-time membership and equivalence query algorithm for exactly learning F implies that the representation problem for F is in fact in co-NP. The theorem is applied to prove hardness results such as the ones mentioned above, by showing that the representation problem for specific classes of formulas is NP-hard. Received: December 6, 1994     

View updates in stratified disjunctive databases

The view update problem is considered in the context of deductive databases where the update of an intensional predicate is accomplished by modifying appropriately the underlying relations in the extensional database. Two classes of disjunctive databases are considered. The first class contains those disjunctive databases which allow only definite rules in the intensional database and disjunctive facts in the extensional database. The second class contains stratified disjunctive databases so that in addition to the first class, negation is allowed in the bodies of the rules, but the database must be stratified. Algorithms are given both for the insertion of an intensional predicate into and the deletion of an intensional predicate from the database. The algorithms use SLD resolution and the concept of minimal models of the extensional database. The algorithms are proved to be correct and best according to the criterion of causing minimal change to the database, where we give first priority to minimizing deletions.Research supported by the National Science Foundation under grant numbers IRI-8916059, IRI-8921591, IRI-9200898, and IRI-9210220.