This is the second in the series of the articles on an application of the systems analytic approach to evaluation of information retrieval (IR). In the previous article a historical overview of IR was presented and existing terminological problems ...
Large amounts of image data have been produced on the Internet over the past several years. As a kind of effective retrieval way, the content-based image retrieval (CBIR) has attracted more and more attention. To improve the preciseness, most CBIR ...
Shared evaluation tasks have become popular over the last decades as ways of making communities of researchers advance together. This paper presents the organization of five new shared task evaluation campaigns for image indexing and retrieval. We have ...
Ian Ruthven
Rosemann and Green consider some ontological constructs defined by Bunge, and made more specific in the context of information systems by Wand and Weber. These constructs provide a semantically clear and solid foundation for understanding and creating domain models of various information systems (IS)-related domains, independently of whether computer-based systems are, or are planned to be, used for automation of some processes within these domains. The authors claim that the understandability of these ontological constructs could be improved by presenting "a meta model [...] using a meta language that is familiar to many IS professionals." The language chosen in the paper is that of an extended entity-relationship (ER) model. These claims are dubious since the semantics of important constructs used in the extended ER model have not been clearly specified (often relying on the fallacy of so-called meaningful names). Second, the representation used in the ER diagrams is not based on semiotic principles (similar, but somewhat different constructs, such as relationships, have not been presented in similar, but somewhat different ways). Third, at least some ER diagrams shown in the paper are unclear. The visual familiarity of ER diagramming constructs may provide an illusion of understandability to some readers of these diagrams, thus giving "a superficial but false sense of security" (Dijkstra). Furthermore, the composition relationship (perhaps the most important one among the ontological constructs by Bunge, Wand, and Weber) has not been properly addressed in the extended ER model. A substantially better treatment of this (and other) ontological constructs was provided by Wand in "A proposal for a formal model of objects" [1], which was not included in the paper's lengthy reference list. In addition, the semantics of important constructs often used in various extended ER models have been clearly defined in international standards, such as the Reference Model of Open Distributed Processing (RMODP) and the General Relationship Model (GRM), and these definitions (also not mentioned in this paper) are close to the ones used by Bunge, Wand, and Weber. Online Computing Reviews Service
Become a reviewer for Computing Reviews.
Morgan Kaufmann Publishers Inc.
San Francisco, CA, United States