Description logics and reasoners, which are descendants of the kl-one language, have been studied in depth in Artificial Intelligence. After a brief introduction, we survey in this paper their application to the problems of information management, using the framework of an abstract information server equipped with several operations -- each involving one or more languages. Specifically, we indicate how one can achieve enhanced access to data and knowledge by using descriptions in languages for schema design and integration, queries, answers, updates, rules, and constraints.

It is natural to view concept and role definitions in Description Logics as expressing monadic and dyadic predicates in Predicate Calculus. We show that the descriptions built using the constructors usually considered in the DL literature are characterized exactly as the predicates definable by formulas in ¨L³, the subset of First Order Predicate Calculus with monadic and dyadic predicates which allows only three variable symbols. In order to handle “number bounds”, we allow numeric quantifiers, and for transitive closure of roles we use infinitary disjunction. Using previous results in the literature concerning languages with limited numbers of variables, we get as corollaries the existence of formulae of FOPC which cannot be expressed as descriptions. We also show that by omitting role composition, descriptions express exactly the formulae in ¨L², which is known to be decidable.

. We describe CARIN, a novel family of representation languages, which integrate the expressive power of Horn rules and of description logics. We address the key issue in designing such a language, namely, providing a sound and complete inference procedure. We identify existential entailment as a core problem in reasoning in CARIN, and describe an existential entailment algorithm for CARIN languages whose description logic component is ALCNR. This algorithm entails several important results for reasoning in CARIN, most notably: (1) a sound and complete inference procedure for non recursive CARIN-ALCNR, and (2) an algorithm for determining rule subsumption over ALCNR. 1 Introduction Horn rule languages have formed the basis for many Artificial Intelligence application languages because their expressive power is sufficient for many applications, and they have good computational properties. One of the significant limitations of Horn rules is that they are not expressive enough to mod...

Knowledge bases, data bases and object-oriented systems (referred to in the paper as Object-Centered systems) all rely on attributes as the main construct used to associate properties to objects; among these, a fundamental role is played by the so-called part-whole relation. The representation of such a structural information usually requires a particular semantics together with specialized inference and update mechanisms, but rarely do current modeling formalisms and methodologies give it a specific "first-class" dignity. The main thesis of this paper is that the part-whole relation cannot simply be considered as an ordinary attribute, its specific ontological nature requires to be understood and integrated within data modeling formalisms and methodologies. On the basis of such an ontological perspective, we survey the conceptual modeling issues involving part-whole relations, and the various modeling frameworks provided by knowledge representation and object-oriented formalisms.

Abstract. One distinctive characteristic of object-oriented data models over traditional database systems is that they provide more expressive power in schema de nition. Nevertheless, the de ning power of objectoriented models is still somewhat limited, mainly because it is commonly accepted that part of the semantics of the application can be represented within methods. The research work reported in this paper explores the possibility of enhancing the power of object-oriented data models in schema de nition, thus o ering more possibilities to reason about the intension of the database and better supporting data management. We demonstrate our approach by presenting a new data model, called CVL, that extends the usual object-oriented data models with several aspects, including view de nition, recursive structure modeling, navigation of the schema through forward and backward traversal of links (attributes and relations), subsetting of attributes, and cardinality ratio constraints on links. CVL is equipped with sound, complete, and terminating inference procedures, that allow various forms of reasoning to be carried out on the intensional level of the database. 1

The aim of the crack project is the research and the development of a knowledge representation architecture based on description logics. The crack system is different from other knowledge representation systems for the high expressivity of the language and the possibility of having sound and complete reasoning procedures. With respect to other systems available in the research community, crack is more expressive, it is expandable to new constructs, it treats the conceptual and individual levels in a homogeneous way, it is modular, and it is comparably fast. However, crack algorithms are not optimal in the worst cases, e.g. in some (arguably rare in practice) worst cases they may require exponential memory. The performance of the system has been tested against several different classes of random knowledge bases, characterized by an order parameter generating phase transitions in the satisfiability probability space.

A data warehouse is a redundant collection of data replicated from several possibly distributed and loosely coupled source databases, organized to answer OLAP queries. Relational views are used both as a speci cation technique and as an execution plan for the derivation of the warehouse data. In this position paper, we summarize the versatility of relational views and their potential. 1

Abstract. The computational complexity of reasoning on pure concept expressions has been characterized completely for all relevant description logics. On the contrary, reasoning in the presence of schema axioms is not so well understood and far from being settled completely. An important class of schemata is that of primitive schemata (in which the schema axioms express only necessary conditions) possibly containing cycles. In this paper we provide, for a relevant class of description logics, a complete characterization of computational complexity of reasoning in these types of schemata, both in the presence and in the absence of cycles. The results are obtained by devising reasoning procedures, establishing direct reductions to show lower bounds, and introducing a general technique by which the constructor for existential quantification can be removed without influencing the result of reasoning. 1

this paper is to rectify this problem, by integrating aspects of the research of [BW94] into DLs. We point out that the success of this integration is not a priori expected because their data model does not involve concept definitions, which would correspond to views in OODB, yet the main concern of DLs is exactly with reasoning about such view definitions. 1.1 Description Logics DLs are used to describe situations using various kinds of individuals (objects), related by relationships, which we call here attributes, and grouped into sets (classes/concepts)

Frame-based knowledge representation and reasoning systems typically provide procedural interfaces for asking about properties of individuals and concepts. We propose an alternative declarative approach that extends standard interface functionality by supporting selective viewing of components of complex objects. Instead of just returning sets of individuals, our queries match concepts and ltered fragments of descriptions. The query language is an extended form of the language used to describe the knowledge-base contents, thus facilitating user training. In this paper, we describe a variety of possible semantics for answering queries in description logics. We investigate the algorithms required when answers are deduced by matching queries against a \structural normal form" of descriptions. As part of our approach, we introduce a useful re nement of the notion of structural subsumption. 1