The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
We present in this paper a test set induction procedure which is refutationally complete for conditional specifications (not restricted to Boolean specifications), in that it refutes any conjecture which is not an inductive theorem. Previously, we could only compute a test set for a conditional specification if the constructors were free. Here, we give a new definition of test sets and a procedure...
We present a simple learning algorithm for equational reasoning. The Knuth-Bendix algorithm can produce deductive consequences from sets of function equations but cannot deduce anything from grounded equations alone. This motivates an inductive procedure which conjectures function equations from a given database of grounded equations.
Cost-sensitive specialization is a generic technique for mis-classification cost sensitive induction. This technique involves specializing aspects of a classifier associated with high misclassification costs and generalizing those associated with low misclassification costs. It is widely applicable and simple to implement. It could be used to augment the effect of standard cost-sensitive induction...
We present a multiple predicate learner (MPL-Core) which efficiently induces some Horn clauses from example sets of multiple predicates and relative background knowledge. Core, a single predicate learning module, has a fast failure mechanism, and can select refinement operators based on the learning task. By means of GPC, an efficient pruning method, Core effectively prunes unpromising branches in...
The increasing availability of a large number of interactive multi-media information services means that users have a large and diverse collection of choices open to them. One method of assisting users to navigate through this large collection is to use information filtering to extract only the information relevant to an end-user according to his/her long-term preferences. In this paper, we describe...
This paper presents work on an interactive fault diagnosis expert system for telecommunication applications. A new knowledge representation and inference algorithm is proposed to suit the characteristics of the application environment, namely: (1) no parallel event exists in human fault reporting, (2) the diagnostic sequence is unpredictable, and (3) the inference engine is passive in an event-driven...
This paper discusses, Galileo, an Intelligent Education System which facilitates learners' scientific thinking. In the discussion we will focus on the concepts behind the designing of Galileo, and the guidance that it provides. First, we will discuss how to support scientific thinking. For this purpose, we will discuss naive knowledge which has been formed through learner's daily experiences. A learner...
In this paper, we present a unified approach to handling uncertainty during plan inference in cooperative consultations. This approach assists with the following aspects of the plan inference process: inferring a user's intentions among a number of possibilities, deciding whether to admit an unlikely interpretation of the user's request or to actively acquire information from the user, determining...
In order to provide a general framework for deductive object-oriented representation systems, Akama's declarative program theory is extended under the assumptions that there exists implicit implication among elements of an interpretation domain and that this implicit implication can be represented by a preorder on the domain. Under the consequent constraint that every interpretation must conform to...
Items are a uniform framework for modelling the data, information and knowledge in a knowledge-based system. If two items share an unstated sub-item between them then any changes to that unstated sub-item will require that both of those two items should also be modified; this situation constitutes a maintenance hazard. Knowledge decomposition reduces each item to a form in which it contains no sub-items...
Although mathematically elegant, partial orders limit the representation of taxonomic knowledge to subsort-supersort (or isa) relationships. We cannot, for example, directly state that two sorts are incompatible or define one sort as the intersection of a set of other sorts. This poses problems for specifying more complete taxonomic relationships as well as for denotational semantics in sorted logic...
A sound and complete view update procedure for a probabilistic deductive database is formulated using SLDp derivation trees introduced by Ng & Subrahmanian in [9]. In order to reduce the number of valid translations that can satisfy an update request a preference criteria is proposed. Moreover, we introduce a method called Δ-factor to minimize the change effected by updates in the database.
We present a logical framework for modelling and reasoning about requirements evolution in the construction of information systems. Our framework represents a requirements model as a theory of some nonmonotonic logic, while requirements evolution is modelled as a mapping between such theories, based on the AGM logic of belief change [1]. We demonstrate our ideas by using the THEORIST system for nonmonotonic...
In this paper, we present a Hierarchical Interactive Reasoning approach for hand-written form processing, and describe an application of this approach to a prototype form processing system. A priori domain knowledge is closely bound to all levels of a hierarchical structure in which bi-directional inferencing is possible. This approach has significantly improved recognition performance in the prototype...
It is reasonable to say that so far neural networks have performed very well on many specific tasks of reasonable size, but their performance is far from satisfactory when applied to realistic but complex tasks such speech recognition and language processing. Yet the brain can perform these tasks efficiently and effortlessly (seemingly) using its optimized mechanisms. It is believed crucial to discover...
A way of processing symbolic information by non-symbol processor is discussed Current computer is designed to process only symbolic information. The first objective of this paper is to discuss that some information is lost by the use of symbolic expression in compensation of its simplicity and this lost part of information often plays an important role in the real world. Some concepts, the detail...
The task is to monitor walking patterns and give early warning of falls using foot switch and mercury trigger sensors. We describe a dynamic belief network model for fall diagnosis which, given evidence from sensor observations, outputs beliefs about the current walking status and makes predictions regarding future falls. The model represents possible sensor error and is parametrised to allow customisation...
In this paper, we present a computational model for transforming discourses into Quasi-Mental Clusters (QMCs) through a convergence process. The process is interpreted as a particular transformation of a given set of discourse segments and concepts by examining the textual continuity. Examinations include testing the local cohesion in a cohesion parsing as well as the global coherence in semantic...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.