Verification Of Knowledge-based Systems Research Articles

An Algebraic Approach to DC Railway Electrification Verification

Grobner bases have been applied to a number of problems related to the verification of Knowledge-Based Systems (KBS) and other problems within graph theory. In particular, the authors have developed in previous papers algebraic approaches to decide whether a situation in a railway interlocking system is safe or not. These algebraic approaches stand out because of the briefness of the code (as they use implementations of well-known algorithms for solving linear or algebraic systems provided by computer algebra systems). The authors have also developed a matrix-based computer tool that helps an expert to check whether a proposed railway electrification scenario (given through the topology of the railway station and the position of the isolation devices and the position and state of the “electrical bypasses” and feeders) fulfills the mandatory requirements of the Spanish railway infrastructure administrator (ADIF) for 3000 V railway electrifications or not. The second author works in the field of railway electrifications and he compares the present-day methods for verification of railway electrifications with the way KBS were verified in the past (manually by experts). In this article we approach this latter problem using algebraic techniques. The new computer tool is based on an algebraic translation of the problem (instead of based on the use of matrices) and is really simple and fast. Determining which electrification sections are under electric tension is computed solving linear systems (because in this case the graph is undirected and no polynomials of degree $$>\,1$$ arise in the algebraic translation, so it is not necessary to use Grobner bases, unlike in the two problems mentioned in the beginning of the Abstract). Therefore far bigger railway facilities can be addressed than if non-linear systems were involved.

Verification of Knowledge-Based Systems Using Predicate/Transition Nets

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> As expert-system technology gains broader acceptance, the need to build and maintain large-scale knowledge-based systems (KBSs) will assume greater importance. Traditional approaches to KBS verification generally contain no predicate/transition (PrT) net models, thus making them slow for the large-scale KBS with chained errors. This paper proposes an attractive alternative to KBS verification, in which the KBS is modeled as a PrT-net model. Then, the least fixpoint semantics of the PrT-net model can be introduced into the KBS for the purpose of speeding up the computations of the KBSs. The significance of this paper is that seven propositions are formulated to detect errors of redundancy, subsumption, unnecessary condition, circularity, inconsistency, dead end, and unreachable goal. Thus, the performance of a computer-aided-design tool for KBSs can be improved to some extent. Meanwhile, specification languages, including Programming in Logic, Frame-and-Rule-Oriented Requirements Specification Language, and the like, are suitable to this approach. </para>

Validation and verification of knowledge-based systems: report on EUROVAV99

Knowledge-Based (KB) technology is being applied to complex problem solving and safety and business critical tasks in many application domains. Concerns have naturally arisen as to the dependability of Knowledge-Based Systems (KBS). As with any software, attention to quality and safety must be paid throughout development of a KBS, and rigorous Verification and Validation (V&amp;V) techniques must be employed. Research in V&amp;V of KBSs has emerged as a distinct field only in the last decade, and is intended to address issues associated with quality and safety aspects of KBSs, and to provide such applications with the same degree of dependability as conventional applications. In recent years, V&amp;V of KBSs has been the topic of annual workshops associated with the main AI conferences, such as AAAI, IJCAI and ECAI.

A Fuzzy Petri Net Tool For Modeling and Verification of Knowledge-Based Systems

A Fuzzy Petri Net Tool For Modeling and Verification of Knowledge-Based Systems

Model-based verification of knowledge-based systems: A case study

As knowledge-based systems become a standard in software development, the interest in verification and validation techniques to ensure their reliability has grown. For this purpose the verification of knowledge bases is fundamental. Numerous references in the literature involve verification by detection of implementation-dependent anomalies in the knowledge base such as inconsistency, incompleteness and redundancy. This approach has several limitations. The main one is the lack of significance of these anomalies in relation to the task of the system. To overcome this problem an approach that exploits the information on the organisation of the knowledge involved in the task, and of the utilisation of this knowledge for solving the task is proposed. This approach, referred to as model-based verification is described, and a case study on the verification of classical planning systems is presented.

A software architecture for knowledge-based systems

The paper introduces a software architecture for the specification and verification of knowledge-based systems combining conceptual and formal techniques. Our focus is component-based specification enabling their reuse. We identify four elements of the specification of a knowledge-based system: a task definition, a problem-solving method, a domain model, and an adaptor. We present algebraic specifications and a variant of dynamic logic as formal means to specify and verify these different elements. As a consequence of our architecture we can decompose the overall specification and verification task of the knowledge-based systems into subtasks. We identify different subcomponents for specification and different proof obligations for verification. The use of the architecture in specification and verification improves understandability and reduces the effort for both activities. In addition, its decomposition and modularisation enables reuse of components and proofs. Therefore, a knowledge-based system can be built by combining and adapting different reusable components.

Using KBS verification techniques to demonstrate the existence of rule anomalies in ADBs

As the field of verification and validation for knowledge-based systems (KBSs) has matured, much information, technology, and theory has become available. Though not all of the problems with respect to KBSs have been solved, many have been identified with solutions that can be used in an analogous manner in situations where the application is not necessarily a traditional KBS. As one example, the “active” component in an active database (ADB) consists of rules that execute as a result of database accesses and updates. In this paper, we demonstrate that anomalies found to impact the correctness of a KBS can also exist in ADBs. We first compare the rule structure of a KBS with the rule structures of various ADBs. To show their existence, we convert the rule syntax of the ADBs into a consistent format for analysis and anomaly detection. Once converted, we apply KBS verification techniques to isolate these anomalies. Due to the more increasing use of triggered rules in ADBs, this work illustrates the danger these anomalies can pose and the ever increasing need for ADB verification techniques to exist.

Verification and Validation of Knowledge-Based Systems: Report on Two 1997 Events

This article gives an overview of two recent events on the validation and verification of knowledge-based systems: (1) the 1997 European Symposium on the Verification and Validation of Knowledge-Based Systems (EUROVAV-97) and (2) the Four-teenth National Conference on Artificial Intelligence Workshop on the Verification and Validation of Knowledge- Based Systems. To give an integrated view of current research issues in this field, we organized this article along thematic lines, unifying the reports of the two separate meetings. Our report focuses on the trends that we think will be important in the near future in this field.

A survey of tools for the validation and verification of knowledge-based systems: 1985–1995

This paper presents the findings of a survey of software tools built to assist in the verification and validation of knowledge-based systems. The tools were identified from literature sources from the period 1985–1995. The tool builders were contacted and asked to complete and return a survey that identified which testing and analysis techniques were utilised and covered by their tool. From these survey results it is possible to identify trends in tool development, technique coverage and areas for future research.

Applying rule-base anomalies to KADS inference structures

The literature on validation and verification of knowledge-based systems contains a catalogue of anomalies for knowledge-based systems, such as redundant, contradictory or deficient knowledge. Detecting such anomalies is a method for verifying knowledge-based systems. Unfortunately, the traditional formulation of the anomalies in the literature is very specific to a rule-based knowledge representation, which greatly restricts their applicability. In this paper, we show how the traditional anomalies can be reinterpreted in terms of conceptual models (in particular KADS inference structures). For this purpose, we present a formalisation of KADS inference structures which enables us to apply the traditional rule-base anomalies to these inference structures. This greatly improves the usefulness of the anomalies, since they can now be applied to a much wider class of knowledge-based systems. Besides this reformulation and wider applicability of the traditional anomalies, further contributions of this paper are a novel formalisation of KADS inference structures and a number of improvements to the existing formalisation of the traditional anomalies.

Evaluation of verification tools for knowledge-based systems

Validation has emerged as a significant problem in the development of knowledge based systems (KBS). Verification of KBS correctness and completeness has been cited as one of the most difficult aspects of validation. A number of software tools have been developed to perform such verification, but none of these are in widespread use. One of the reasons for this is that little quantitative evidence exists to demonstrate the effectiveness of the tools. This paper presents an experimental study of three KBS verification tools: a consistency checker, a completeness checker and a testing tool (for correctness). The tools are evaluated on their ability to reveal plausible faults seeded into a complex, realistic KBS application. The cost of using the tools is also measured. It is shown that each tool is independently effective at detecting certain kinds of fault and that the capabilities of the tools are complementary—a result not revealed by previous studies.

ELIMINATING REDUNDANCY, CONFLICT, AND INCOMPLETENESS FROM KNOWLEDGE-BASED SYSTEMS

The reliability of knowledge-based systems (KBSs) has been the subject of a great deal of recent research. Much of this research focuses on the verification of KBSs, specifically to eliminate redundant rules, conflicting rules, and to ensure that the KBS is complete. Often, verification is approached by testing the structural properties of the KBS after the rules have been defined. In this paper we take a different approach and show how the process of formal program derivation from software engineering can be applied to KBSs. The use of these techniques eliminates the need for post-development verification because program derivation guarantees that the KBS will not contain redundant rules, conflicting rules, or be incomplete. As such, verification concerns are addressed during development.

Eighth Workshop on the Validation and Verification of Knowledge-Based Systems

The Workshop on the Validation and Verification of Knowledge-Based Systems gathers researchers from government, industry, and academia to present the most recent information about this important development aspect of knowledge-based systems (KBSs). The 1995 workshop focused on nontraditional KBSs that are developed using more than just the simple rule-based paradigm. This new focus showed how researchers are adjusting to the shift in KBS technology from stand-alone rule-based expert systems to embedded systems that use object-oriented technology, uncertainty, and nonmonotonic reasoning.

The Seventh Workshop on the Validation and Verification of Knowledge-Based Systems

The annual Workshop on the Validation and Verification of Knowledge-Based Systems is the leading forum for presenting research on the validation and verification of knowledge-based systems (KBSs). The 1994 workshop was significant in that there was a definitive move in the philosophical position of the workshop from a testing- and tool-based approach to KBS evaluation to that of a formal specification-based approach. This workshop included 12 full papers and 5 short papers and was attended by 35 researchers from government, industry, and academia.

Knowledge-based systems verification: A machine learning-based approach

This paper addresses the problem of verification of knowledge bases. It presents a knowledge-based system (KBS) verification approach that considers system specifications and, consequently, knowledge bases to be partially described when development starts. This partial description is not necessarily perfect, and our work aims at using machine learning techniques to progressively acquire new knowledge and then improve the quality of expert system knowledge bases (KB) by coping with two major KB anomalies: incompleteness and incorrectness. The KBs considered in our approach are expressed in different formalisms.

A DEVELOPMENT FRAMEWORK AND VERIFICATION METHODOLOGIES FOR KNOWLEDGE-BASED SYSTEMS

In this paper, we present a development framework and verification methodologies for knowledge-based systems (KBSs) with real-time systems as our target system. The framework originates from an integration of three software development paradigms: rapid prototyping, operational specification, and transformational implementation. Based on this framework, we present RT-FRORL as a formal requirements specification language which exploits knowledge representation techniques as an aid in the specification, development, and verification of a KBS for real-time systems. RT-FRORL uses a combination of resolution refutation, anomaly detection matrix, and algorithms methods to verify a number of properties which might exist in KBSs. By incorporating RT-FROHL and its verification methods with the underlying framework, it lays a very strong foundation to deal with the current issues in KBSs verification.

A formal model for verification of dynamic consistency of KBSs

This paper proposes a translation of the main concepts involved in Knowledge Based Systems Verification into a theoretical metalanguage based on Halmos and Leblanc's “Monadic and Polyadic Algebras.” These algebras are expressed in terms of a few basic concepts of preorder-category theory. Any Knowledge Base (KB) may be considered as a set of arrows that gives rise to a preorder category C, called “the N-category associated to the KB.” Two subsets are defined in C: E q = { x : x → q is an arrow in C}, and E p = { x : p → x is an arrow in C}, where q and p, respectively, are a goal and a conjunction of facts. A logic is a pair ( C, E p ); it is consistent if it is not the case that both a proposition and its negation are simultaneously in E p , which is equivalent to the inequality E p ≠ C. In this context, the two main ideas in the paper are the following. First, to characterize forward reasoning consistency of a KB with respect to a set of facts, in terms of consistency in ( C, E p ), where C is the N-category associated to the KB, and p is the conjunction of all the facts in the given set. Second, to characterize the absence of conflict in backward reasoning in terms of some relations among sets of the form E p and E q , for appropriate p and q. The aim of the paper is to present ideas for further discussion leading to the construction of formal logico-algebraic models for verification, rather than to propose a completed model. It presents just one possible approach which may suggest others.

AAAI-93 Workshops: Summary Reports

The Association for the Advancement of Artificial Intelligence sponsored a number of workshops in conjunction with the Eleventh National Conference on Artificial Intelligence held 11-15 July 1993 in Washington, D.C. This article contains reports of four of the workshops that were conducted: AI Models for System Engineering, Case-Based Reasoning, Reasoning about Function, and Validation and Verification of Knowledge Based Systems.

A formal model for knowledge-based systems verification

A logico-algebraic model for verification of Knowledge Based Systems, based on Abraham Robinson's meta-algebra, is presented in this article. the set of rules of Knowledge Based System is considered to be the set of axioms of a theory “T”. the model consists of: (i) the collection “P” of all well formed formulas of the language of T, (ii) a subset P0 of P that has as its elements the axioms of a given logic (bivalued, intuitionistic, or other), plus all the theorems that can be deduced from this logic inside the language of T, (iii) some distinguished subsets of P, to be called m-ideals and m-filters, that depend on T and P0. Important concepts that are involved in the verification of Knowledge Based Systems are considered, such as forward and backward reasoning consistencies. Appropriate characterizations are obtained by using properties of m-ideals and m-filters. © 1994 John Wiley & Sons, Inc.

Validation and verification of knowledge-based systems: A survey

Validation and verification of expert systems or knowledge-based systems is a critical issue in the development and deployment of robust systems. This article is a comprehensive survey of the developments and trends in this field. More than 300 references are included in the References and Additional Readings at the end of article.