Research on ontology is becoming increasingly widespread in the computer science community, and its importance is being recognized in a multiplicity of research fields and application areas, including knowledge engineering, database design and integration, information retrieval and extraction. We shall use the generic term information systems, in its broadest sense, to collectively refer to these application perspectives. We argue in this paper that so-called ontologies present their own methodological and architectural peculiarities: on the methodological side, their main peculiarity is the adoption of a highly interdisciplinary approach, while on the architectural side the most interesting aspect is the centrality of the role they can play in an information system, leading to the perspective of ontology-driven information systems.

The purpose of this paper is to explore some semantic problems related to the use of linguistic ontologies in information systems, and to suggest some organizing principles aimed t o solve such problems. The taxonomic structure of current ontologies is unfortunately quite complicated and hard to understand, especially for what concerns the upper levels. I will focus here on the problem of ISA overloading, which I believe is the main responsible of these difficulties. To this purpose, I will carefully analyze the ontological nature of the categories used in current upper-level structures, considering the necessity of splitting them according to more subtle distinctions or the opportunity of excluding them because of their limited organizational role.

Current ontologies' upper-level taxonomic structure is often quite complicated and hard to understand. In this paper I show how the theoretical tools of so-called Formal Ontology, and especially the theory of identity, can help to formulate cleaner, more general, more rigorous, and more understandable upper-level ontologies. I focus in particular on some examples of multiple generalization, proposing a way of simplifying the domain structure by splitting some concepts according to different identity conditions, or by excluding them because of their limited organizational role.

This paper deals with the ontology for natural language understanding. In particular, the paper focuses on the WordNet noun ontology, proposes some additions, changes and reorganizations, and explains the rationale supporting them. These changes have been pointed out by a semantic interpreter driven by over 3000 verb predicates that use WordNet noun ontology and WordNet verb classes. The selectional restrictions in the semantic roles of the predicates are WordNet ontological categories. The main criterion for changing an ontological category has been failing to interpret some sentences. A detailed analysis of the proposed changes is presented. 1

The upper-level taxonomic structure of current ontologies is often quite complicated and hard to understand. In this paper I analyze the ontological nature of upper-level categories, considering the necessity of splitting them according to more subtle distinctions or the opportunity of excluding them because of their limited organizational role. 1. Introduction Currently, a number of efforts in the area of knowledge and language engineering are aimed to the development of systems of basic semantic categories (often called "upper-level ontologies"), to be used for various applications such as natural-language processing, multilingual information retrieval, information integration, intelligent systems design. Examples of such systems are the CYC upper level, the Penman Upper Model [Bateman et al. 1990], recently evolved into the Pangloss ontology [Knight and Luk 1994] and the Revised Upper Model [Bateman 1995], the Mikrokosmos ontology [Mahesh 1996], and the Wordnet upper structure [M...