Abstract: The content of real-world databases, knowledge bases, database models, and formal specifications is often highly redundant and needs to be aggregated before these representations can be successfully paraphrased into natural language. To generate natural language from these representations, a number of processes must be carried out, one of which is sentence planning where the task of aggregation is carried out. Aggregation, which has been called ellipsis or coordination in Linguistics, is the process that removes redundancies during generation of a natural language discourse, without losing any information. The article describes a set of corpus studies that focus on aggregation, provides a set of aggregation rules, and finally, shows how these rules are implemented in a couple of prototype systems. We develop further the concept of aggregation and discuss it in connection with the growing literature on the subject. This work offers a new tool for the sentence planning phase of natural language generation systems.

Abstract: Formal Concept Analysis is a symbolic learning technique derived from mathematical algebra and order theory. The technique has been applied to a broad range of knowledge representation and exploration tasks in a number of domains. Most recorded applications of Formal Concept Analysis deal with a small number of objects and attributes, in which case the complexity of the algorithms used for indexing and retrieving data is not a significant issue. However, when Formal Concept Analysis is applied to exploration of a large numbers of objects and attributes, the size of the data makes issues of complexity and scalability crucial. This paper presents the results of experiments carried out with a set of 4,000 medical discharge summaries in which were recognized 1,962 attributes from the Unified Medical Language System (UMLS). In this domain, the objects are medical documents (4,000) and the attributes are UMLS terms extracted from the documents (1,962). When Formal Concept Analysis is used to iteratively analyze and visualize these data, complexity and scalability become critically important. Although the amount of data used in this experiment is small compared with the size of primary memory in modern computers, the results are still important because the probability distributions that determine the efficiencies are likely to remain stable as the size of the data is increased. Our work presents two outcomes. First, we present a methodology for exploring knowledge in text documents using Formal Concept Analysis by employing conceptual scales created as the result of direct manipulation of a line diagram. The conceptual scales lead to small derived purified contexts that are represented using nested line diagrams. Second, we present an algorithm for the fast determination of purified contexts from compressed representation of the large formal context. Our work draws on existing encoding and compression techniques to show how rudimentary data analysis can lead to substantial efficiency improvements in knowledge visualization.

Abstract: Braille, a touch-reading system for visually impaired people was first introduced in 1825 by Louis Braille. Editing and reprinting Braille text that was originally embossed on paper is both time consuming and labour intensive. This paper presents an automatic system to recognize Braille pages and to convert Braille documents into English or Chinese text for editing.

Abstract: A natural language collaborative consultation system must take user preferences into account. A model of user preferences allows a system to appropriately evaluate alternatives using criteria of importance to the user. Additionally, decision research suggests both that an accurate model of user preferences could enable the system to improve a user's decision-making by ensuring that all important alternatives are considered, and that such a model of user preferences must be built dynamically by observing the user's actions during the decision-making process. This paper presents two strategies: one for dynamically recognizing user preferences during the course of a collaborative planning dialogue and the other for exploiting the model of user preferences to detect suboptimal solutions and suggest better alternatives. Our recognition strategy utilizes not only the utterances themselves but also characteristics of the dialogue in developing a model of user preferences. Our generation strategy takes into account both the strength of a preference and the closeness of a potential match in evaluating actions in the user's plan and suggesting better alternatives. By modeling and utilizing user preferences, our system is able to fulfill its role as a collaborative agent.

Abstract: With the increasing size of databases, the extraction of data summaries becomes more and more useful The use of fazzy sets seems interesting in order to extract linguistic summaries, ie, statements from the natural language, containing gradual properties, which are meaningful for human operators This paper focuses on the extraction from databases of linguistic summaries, using so-called fuzzy gradual rules, which encode statements of the form "the younger the employees, the smaller their bonus" The summaries considered here are more on the relations between labels of the attributes than on the data themselves The first idea is to extract all the rules which are not in contradiction with tuples of a given relation Then, the interest of these rules is questioned For instance, some of them can reveal potential incoherence, while other are not really informative It is then shown that in some cases, interesting information can be extracted from these rules Last, some properties the final set of rules should verify are outlined

Abstract: Efficient implementation of type inclusion is an important feature of object oriented programming languages with multiple inheritance. The idea is to associate to each type a subset of a set S={1,...,k} such that type inclusion coincides with subset inclusion. Such an embedding of types into 2S (the lattice of all subsets of S) is called a bit-vector encoding of the type hierarchy. In this paper, we show that most known bit-vector encoding methods can be inserted on a general theoretical framework using graph coloration, namely the notion of a simple encoding. We use the word simple because all these methods are heuristics for the general bit-vector encoding problem, known as the 2-dimension problem. First we provide a correct algorithm for partial orders based on simple encoding, improving the algorithm of Krall, Vitek, and Horspool (1997). Second we show that finding an optimal simple encoding is an NP-hard problem. We end with a discussion on some practical issues.

Abstract: We develop a genetic algorithm as a method for cryptanalysing the Chor-Rivest knapsack public key cryptosystem (PKC) (B. Chor and R.L. Rivest, 1988). As far as we know there is no feasible attack known on it (A.J. Menezes, 1997). The results show how the algorithm is effectively used to break this scheme by examining a very small fraction of the space of possible solutions. The algorithm found the exact solution in all attempted cases.

Abstract: We discuss the conjugacy problem in the family of fuzzy implications. Particularly we examine a compatibility of conjugacy classes with induced order and induced convergence in the family of fuzzy implications. Conjugacy classes can be indexed by elements of adequate groups.

Abstract: In recent years, radial basis function (RBF) neural networks have been shown to be versatile for performance prediction involving nonlinear processes. Machining performance prediction involving various process variables is a nonlinear problem. The conventional mechanics of the cutting approach for predicting thrust and torque in drilling makes use of the oblique cutting theory and an orthogonal cutting databank. The quantitative reliability, in these models, depends on the 'input parameters' along with the 'edge force' components from the orthogonal cutting databank for that given work material. By contrast, neural networks for drilling performance prediction have been shown to be successful for quantitative predictions with minimum number of inputs. In this paper, an RBF neural network architecture is proposed which uses process variables such as tool geometry and operating conditions to estimate thrust and torque in drilling. Extensive drilling tests are carried out to train the RBF network. The developed network is tested over a range of process variables to estimate thrust and torque. It is shown that, using the neural network architecture, the drilling forces are 'simultaneously' predicted to within 5% of the experimental values.

Abstract: Many optimization problems consist of several mutually dependent subproblems, where the resulting solutions must satisfy all requirements. We propose a new model for Multi-Objective Optimization (MOO) in Evolutionary Algorithms (EAs). The search space is partitioned into so-called Satisfiability Classes fSCj, where each region represents the quality of the optimization criteria. Applying the SCs to individuals in a population a fitness can be assigned during the EA run. The model also allows the handling of infeasible regions and restrictions in the search space. Additionally, different priorities for optimization objectives can be modeled. Advantages of the model over previous approaches are discussed and an application is given that shows the superiority of the method for modeling MOO problems.

Abstract: A critical problem for managers of temporal information is the treatment of assertions and of complex types of queries because in many cases the treatment could involve reasoning on the whole knowledge base of temporal constraints. We propose an efficient approach to this problem. First, we show how different types of queries can be answered (in a complete way) in a time polynomial in the dimension of the query and independently of the dimension of the knowledge base. Second, we provide an efficient (and complete) procedure to deal with sessions of interleaved assertions and queries to the knowledge base. We provide both analytical and experimental evaluations of our approach, and we discuss some application areas.

Abstract: The clinical interpretation of fetal heart rate traces is a difficult task that has led to the development computerised assessment systems. These systems are limited by their inability to represent uncertainty. This paper describes the first stage in the development of a fuzzy expert system for fetal heart rate assessment. A preliminary evaluation study comparing the initial fuzzy system with three clinicians and an existing crisp expert system is presented. The fuzzy system improved on the crisp system and achieved the highest overall performance. The use of fuzzy systems for analysis of fetal heart rate traces and similar time varying signals is shown to have potential benefit.

Abstract: This paper shows how a multi-agent systems (MAS) approach may be used to build an interactive intelligent tutoring system (ITS) designed as a game. We model the ITS as a society of both reactive and cognitive agents that interact through a graphical interface. We present an experiment carried out in a school in order to test our hypothesis about the architecture's pedagogical potential, and we present the results obtained.

Abstract: The finishing train of a hot strip mill has been modelled by using a constant volume element model. The accuracy of the model has been increased by using an Artificial Neural Network (ANN). A non-linear Rank Based Geaetic Algorithm has been developed for the optimization of the work roll profiles in the finishing stands of the simulated hot strip mill. It has been compared with eight other experimental optimization algorithms: Random Walk, Hill Climbing, Simulated Annealing (SA) and five different Genetic Algorithms (GA). Finally, the work roll profiles have been optimized by the non-linear Rank Based Genetic Algorithm. The quality of the strip from the simulated mill was significantly improved.

Abstract: We propose an interactive approach to the linguistic summarization of databases. The point of departure is that a fully automatic generation of summaries is presently impossible, and an interaction with the user is necessary; the user is to specify a class of summaries of interest or relevance. For an implementation of such an interactive approach, a human-friendly database querying interface is employed, and we propose to use our FQUERY for Access querying add-on. An implementation of two relevant types of summaries is shown.

Abstract: We present a method to derive a solution to the combined frame and ramification problems for certain classes of theories of action written in the situation calculus. The theories of action considered include the causal laws of the domain, in the form of a set of effect axioms, as well as a set of ramification state constraints. The causal laws state the direct effects that actions have on the world, and ramification state constraints allow one to derive indirect effects of actions on the domain. To solve the combined frame and ramification problems, the causal laws and ramification state constraints are replaced by a set of successor state axioms. Given a state of the world, these axioms uniquely determine the truth value of dynamic properties after an action is performed. In this article, we extend previous work by formulating an approach for the mechanical generation of these successor state axioms. We make use of the notions of implicate and support that have been developed in the context of propositional theories. The approach works for classes of syntactically restricted sets of ramification state constraints.

Abstract: Fuzzy set theory provides an approximate but effective means of describing the behavior of ill-defined systems. Patterns of human origin such as handwritten characters are to some extent found to be fuzzy in nature. The authors decided to use the fuzzy conceptual approach. The paper attempts to use the fuzzy concept on handwritten Tamil characters to classify them as one among the prototype characters using a feature called distance from the frame and a suitable membership function. The prototype characters are categorized into two classes: one is considered as line characters/patterns and the other as arc patterns. The unknown input character is classified into one of these two classes first and then recognized to be one of the characters in that class. The algorithm is tested for about 250 samples for seven chosen Tamil characters and the success rate obtained varies from 88% to 100%.

Abstract: Multi-valued and universal binary neurons (MVN and UBN) are the neural processing elements with complex-valued weights and high functionality. It is possible to implement an arbitrary mapping described by partial-defined multiple-valued function on the single MVN and an arbitrary mapping described by partial-defined or fully-defined Boolean function (which can be not threshold) on the single UBN. The fast-converged learning algorithms are existing for both types of neurons. Such features of the MVN and UBN may be used for solution of the different kinds of problems. One of the most successful applications of the MVN and UBN is their usage as basic neurons in the Cellular Neural Networks (CNN) for solution of the image processing and image analysis problems. Another effective application of the MVN is their use as the basic neurons in the neural networks oriented to the image recognition.

Abstract: Theoretically, fuzzy control has been shown to be superior in complex problems with multi-objective decisions. Traffic signal control is a typical process, where traffic flows compete from the same time and space, and different objectives can be reached in different traffic situations. Based on recent research work, fuzzy control technology appears particularly well suited to traffic signal control situations involving multiple approaches and vehicle movements. Based on the results of our paper, we can say that the fuzzy control principles are very competitive in isolated multiphase traffic signal control. The experiences and results of the field test and the calibration of membership functions with neural networks have been extremely promising.

Abstract: The article presents a neurofuzzy network that is applied to the detection of a specific wave of the electrocardiographic signal. The network was trained using genetic algorithms, using a software package publicly available on the Internet. The training procedure, its parameters and details of the application are presented. Results suggest that this kind of network is suitable for the identification of patterns in unidimensional time-varying signals.

Abstract: HyperText Markup Language (HTML) is a platform-independent method of identifying a document's structure and references. The Web is a distributed hypertext multimedia information service that has brought millions of non-academic and academic users to the Internet. In constrast to traditional GIS, multimedia Web GIS is not only able to collect, analyze and store data in traditional formats i.e. text, images and graphs but also audio, animations and video, using all the techniques, advantages and multimedia facilities of the Web. This paper explains multimedia data and the advantages of multimedia GIS (MMGIS) over traditional GIS. HTML plays a robust role in the development of the multimedia Web GIS. Through HTML our multimedia Web GIS can be viewed in three different categories. These are Static Web GIS, Dynamic Web GIS, and Active Web GIS. These three broad categories are also discussed.

Abstract: In this paper we propose a new method of tuning the probabilities of the genetic operators. We assume that every member of the optimized population conducts his own ranking of genetic operator qualities. This ranking becomes a base to compute the probabilities of appearance and execution of genetic operators. This set of probabilities is a base of experience of every individual and according to this it chooses the operator in every iteration of the algorithm. Due to this experience one can maximize the chance of offspring survival.

Abstract: In this paper we develop a formalization of semantic relations that facilitates efficient implementations of relations in lexical databases or knowledge representation systems using bases. The formalization of relations is based on a modeling of hierarchical relations in Formal Concept Analysis. Further, relations are analyzed according to Relational Concept Analysis, which allows a representation of semantic relations consisting of relational components and quantificational tags. This representation utilizes mathematical properties of semantic relations. The quantificational tags imply inheritance rules among semantic relations that can be used to check the consistency of relations and to reduce the redundancy in implementations by storing only the basis elements of semantic relations. The research presented in this paper is an example of an application of Relational Concept Analysis to lexical databases and knowledge representation systems (cf. Priss 1996) which is part of a larger framework of research on natural language analysis and formalization.

Abstract: Many artificial and natural systems are often more adequately modelled using data structures, e.g., graphs, trees. For example, it is often more convenient to represent an image using data structures than by representing it using pixels. The data structure can serve as a prelude to scene analysis, image retrieval, etc. There are, broadly speaking, two ways in which data structures can be processed. One way is to consider it as generated by an underlying grammar, with defined syntax; and the other way is to consider it as an input output system, which may be modelled using neural networks. In this talk, we will discuss both approaches. We will first discuss how data structures can be modelled using an attributed grammar. Then, we will discuss the modelling of a data structure using neural networks. It is shown that both approaches are closely related. We will then derive training algorithms for the neural network models, and discuss the universal approximation properties of such models. We will demonstrate the neural network approach on a number of synthesized and practical examples.

Abstract: The intention of this article is to show how fuzzy set theory fits into classical topology. The basic concepts are filter and ideal bases and morphisms on these. Filter bases are used to define abstract distances. Then compositions and comparisons of filter and ideal bases are considered and uniform neighborhood measures between such bases are introduced. On filter bases homomorphisms and antimorphisms can be defmed as set fimctions, in particular, they can be given by the set extension of point functions. With these concepts, fuzzy set theory can be expressed in terms of topology. For some applications we consider contractive mappings, roundings, hierarchical filter bases ("pyramids"), adaptable networks.

Abstract: The problem of inserting a new element x into a lattice of types L is addressed in the paper. As the poset L+x obtained by the direct insertion of x in L is not necessarily a lattice, some set of auxiliary elements should be added to restore the lattice properties. An approach toward the lattice insertion is presented which allows the set of auxiliary elements to be kept minimal. The key idea is to build the final lattice L+ as isomorphic to the Dedekind–McNeille completion of the order L+x. Our strategy is based on a global definition of the set of auxiliary elements and their locations in L+. Each auxiliary is related to a specific element of L, an odd, which represents GLB (LUB) of some elements in L superior (inferior) to x. An appropriate computation scheme for the auxiliary types is given preserving the subtyping in the lattice L+. The insertion strategy presented is more general than the existing ones, since it deals with general kinds of lattices and makes no hypothesis on the location of x in L. An algorithm computing L+ from L and x of time complexity O(|L||J(L)|ω^3(L)) is provided.

Abstract: Multimedia applications within the World Wide Web (WWW) have to deal with difficulties like executing within Web pages and being transferred via the Internet. However, the temporal aspects of hypermedia features for continuous media like audio and video resemble all other kinds of multimedia applications. These temporal aspects are discussed in consideration of presentation and authoring facilities. A system architecture and implementation relying on commercial WWW technology is presented.

Abstract: In this paper we consider two related types of data dependencies that can hold in a relation: conjunctive implication rules between attribute-value pairs, and functional dependencies We present a conceptual clustering approach that can be used, with some small modifications, for inferring a cover for both types of dependencies The approach consists of two steps First, a particular clustered representation of the relation, called concept (or Galois) lattice, is built Then, a cover is extracted from the lattice built in the earlier step Our main emphasis is on the second step We study the computational complexity of the proposed approach and present an experimental comparison with other methods that confirms its validity The results of the experiments show that our algorithm for extracting implication rules from concept lattices clearly outperforms an earlier algorithm, and suggest that the overall lattice-based approach to inferring functional dependencies from relations can be seen as an alternative to traditional methods

Abstract: Learning concepts and rules from structured (complex) objects is a quite challenging but very relevant problem in the area of machine learning and knowledge discovery. In order to take into account and exploit the semantic relationships that hold between atomic components of structured objects, we propose a knowledge discovery process, which starts from a set of complex objects to produce a set of related atomic objects (called contexts). The second step of the process makes use of the concatenation product to get a global context in which binary relations of individual contexts coexist with relations produced by the application of some operators to individual contexts. The last step permits the discovery of concepts and implication rules using the concept lattice as a framework in order to discover and interpret nontrivial concepts and rules that may relate different components of complex objects. This paper focuses on two main steps of the knowledge discovery process, namely data mining and interpretation.